Rep: OEB Doc: 12WVZ Rev: 0 ONTARIO ENERGY BOARD Volume: 5 10 OCTOBER 2003 BEFORE: P. SOMMERVILLE PRESIDING MEMBER A. BIRCHENOUGH MEMBER 1 RP-2003-0063 2 IN THE MATTER OF the Ontario Energy Board Act, 1998, S.O. 1998, c.15 (Sched. B); AND IN THE MATTER OF an Application by Union Gas Limited for an Order or Orders approving or fixing just and reasonable rates and other charges for the sale, distribution, storage, and transmission of gas for the period commencing January 1, 2004. 3 RP-2003-0063 4 10 OCTOBER 2003 5 HEARING HELD AT TORONTO, ONTARIO 6 APPEARANCES 7 PAT MORAN Board Counsel MARTIN DAVIES Board Staff JAMES WIGHTMAN Board Staff MICHAEL PENNY Union Gas Limited TOM BRETT Ontario Association of School Business Officials MICHAEL JANIGAN Vulnerable Energy Consumers' Coalition ROBERT WARREN Consumers Association of Canada ALICK RYDER City of Kitchener DWAYNE QUINN City of Kitchener GEORGE VEGH CEED, OESC, Superior Energy Management, Union Energy, TransAlta JAY SHEPHERD Ontario Public School Boards Association MIMI SINGH CME RANDY AIKEN London Property Management Association, Wholesale Gas Service Purchasers Group SCOTT STOLL Northern Cross Energy TIBOR HAYNAL TransCanada PipeLines ROBERT ROWE Enbridge Gas Distribution Inc. PETER THOMPSON Industrial Gas Users Association BRIAN DINGWALL Energy Probe, HVAC Coalition, Distributed Energy Association DERECK FRANCIS Energy Objective VALERIE YOUNG Ontario Association of Physical Plant Administrators PETER SCULLY City of Timmins, City of Sudbury, FNOM JOHN RATTRAY Ontario Power Generation DAVID BROWN Coral Energy 8 TABLE OF CONTENTS 9 PRELIMINARY MATTERS: [18] UNION GAS LIMITED - PANEL 4; RESUMED; GARDINER, POREDOS, ROGERS [34] CROSS-EXAMINATION BY MR. JANIGAN: [38] CROSS-EXAMINATION BY MR. AIKEN: [628] CROSS-EXAMINATION BY MR. BRETT: [789] MOTION BY CORAL ENERGY: [1018] SUBMISSIONS BY MR. BROWN: [1019] SUBMISSIONS BY MR. PENNY: [1023] RULING: [1031] UNION GAS LIMITED - PANEL 4; RESUMED; GARDINER, POREDOS, ROGERS [1069] CROSS-EXAMINATION BY MR. SHEPHERD: [1073] CROSS-EXAMINATION BY MS. SINGH: [1121] CROSS-EXAMINATION BY MR. DINGWALL: [1166] CROSS-EXAMINATION BY MR. MORAN: [1265] RE-EXAMINATION BY MR. PENNY: [1477] 10 EXHIBITS 11 EXHIBIT NO. M.5.1: MATERIALS FOR CROSS-EXAMINATION FILED BY VECC [77] EXHIBIT NO. M.5.2: SERIES OF GRAPHS EXCERPTED FROM UNDERTAKING J.34.33 [121] EXHIBIT NO. M.5.3: DOCUMENT ENTITLED "LPMA/WGSPG CROSS-EXAMINATION MATERIALS 2" [651] 12 UNDERTAKINGS 13 UNDERTAKING NO. N.5.1: TO PROVIDE EX-POST FORECAST OF THE UNION NAC MODEL [202] UNDERTAKING NO. N.5.2: TO PROVIDE THE NAC TREND FOR THE HISTORICAL PERIOD SHOWN ON FIGURE 2, FIGURE 3, FIGURE 4, AND FIGURE 5, IN EXHIBIT C.1, TAB 1, AT PAGES 11 OF 15 AND FOLLOWING, BY NORMALIZING THE ACTUAL NAC BY THE CONSTANT FORECASTED HDDS FOR 2004, GENERATED FROM THE 20-YEAR TREND METHODOLOGY AND THE 30-YEAR AVERAGE [331] UNDERTAKING NO. N.5.3: TO CONFIRM THAT UNION IS ACCEPTING THE DECREASE IN THE M9 VOLUMES AS REFLECTED IN EXHIBIT K.35 EVIDENCE OF THE WHOLESALE GAS SERVICES PURCHASERS GROUP [778] UNDERTAKING NO. N.5.4: TO PROVIDE THE 2003 CONSUMPTION BY THE 11 NON-UTILITY GENERATORS [807] UNDERTAKING NO. N.5.5: TO TAKE WHAT WAS FORECAST FOR 2004 USING THE 30-YEAR MODEL AND THE HDD FORECAST USING THE 20-YEAR MODEL AND THEN WEIGHT THEM AS THE INPUT FOR THE VOLUME FORECAST [1439] 14 --- Upon commencing at 9:35 a.m. 15 MR. SOMMERVILLE: Good morning. Please be seated. Thank you. 16 Are there any preliminary matters before we commence the cross-examination on this panel? 17 MR. PENNY: Mr. Chairman, from my perspective, just two very briefly. 18 PRELIMINARY MATTERS: 19 MR. PENNY: We will file normal errata on the transcript in the ordinary course, but there were two that I wanted to address as they were potentially matters of substance. The witness, Mr. Poredos and Mr. Rogers, have advised me that there's one for each of them. There's a reference at paragraph 216 of yesterday's transcript from Mr. Poredos in the third last line of that paragraph where it says "those contracts are coming up for renewal." The transcript reads they are "not market." What Mr. Poredos said or intended to say, I think he said it, was "they are at market." So that should read "at market." 20 And then on paragraph 595, an answer of Mr. Rogers, in the second line of that paragraph, it reads "our forecasting information says that it will do" and he either said or intended to say "will not do" which I think you'll see is consistent with the earlier part of the sentence. So there was that. 21 MR. SOMMERVILLE: Thank you. 22 MR. PENNY: And then, on a scheduling matter, it appears as if we have a full day for panel 4 and would not get to panel 5, and on that basis we have advised the gas supply panel they will not be reached today. We had discussed with Mr. Moran the fact that, although we wanted to do gas supply next, the HR group we had to lock in because there are people coming from British Columbia to testify on that panel and so they're going to be here on Tuesday. 23 So I think the best thing to do is just to proceed, given that we've fallen about half a day behind, which isn't bad, I don't mean that, in the scheme of things that's not bad, but I think we'll have to ensure that we get panel 6 on and off on Tuesday and/or Wednesday. We'd better start with panel 6 on Tuesday morning and then we'll come back to gas supply when we're done with them. 24 MR. SOMMERVILLE: Thank you. 25 MR. PENNY: That's all I have. Otherwise, the witnesses are available to continue with the examination. 26 MR. SOMMERVILLE: Thank you. 27 We should try to find some process to ensure that that change in scheduling for those panels is communicated. I know we have a hot line. 28 MR. MORAN: Yes, Mr. Chair, we'll put it on the hot line and we'll send an e-mail as well. 29 MR. SOMMERVILLE: Thank you kindly. 30 Are there any other preliminary matters? I take it we're going to deal with panel 4 today to the conclusion of today's evidence, so anyone who was anticipating cross-examination of panel 5, gas supply, and has no further interest in this panel can safely begin their long weekend now. 31 MR. PENNY: That's correct. 32 MR. SOMMERVILLE: I don't see an exodus at the moment. So without further ado, Mr. Janigan. 33 MR. JANIGAN: Thanks very much, Mr. Chair. 34 UNION GAS LIMITED - PANEL 4; RESUMED; GARDINER, POREDOS, ROGERS 35 P.GARDINER; Previously sworn. 36 S.POREDOS; Previously sworn. 37 B.ROGERS; Previously sworn. 38 CROSS-EXAMINATION BY MR. JANIGAN: 39 MR. JANIGAN: Panel, my name is Michael Janigan. I act on behalf of the Vulnerable Energy Consumers' Coalition. I have some questions in two areas, two broad areas of your evidence. 40 The first deals with your forecasts. And as I understand it, the volume forecast for general service demand for natural gas is arrived at by taking the total number of customers and multiplying it times the normalized average consumption per customer; is that correct? 41 MR. GARDINER: Yes, it is. 42 MR. JANIGAN: Okay. And the normalized average consumption, according to your evidence, is derived from a separate NAC for each of the following rate groups, and that's set out on Exhibit C.1, tab 1, page 4? 43 MR. GARDINER: I agree. 44 MR. JANIGAN: And that's the residential M2, the residential R1, commercial M2, commercial R1, commercial R10, industrial R10, and industrial M2. Am I correct? 45 MR. GARDINER: Yes, you are. 46 MR. JANIGAN: And have the equations that you are using to derive the normalized average consumption been used previously by Union in forecasting NAC? 47 MR. GARDINER: Yes. The equations that are in the evidence we used in the previous budget at Union Gas, and the equation 1, use-per-customer type specification, was used in an earlier budget. So the use equation econometrics has been used at Union Gas for a number of years. The combination of the volume and usage equation approach has been used now for two years at Union Gas. 48 MR. JANIGAN: Now, have you used these equations in a rate case before? 49 MR. GARDINER: Econometric regressions have been used by Union Gas in the past in rate cases. 50 MR. JANIGAN: Have you used these particular equations in rate cases in the past? 51 MR. GARDINER: Not these. In the earlier days, they were similar in some regards, at a very high level, in the sense that they had weather and efficiency in them. The actual specification of the equations may be slightly different, but conceptually they're the same. 52 MR. JANIGAN: And I assume that Union proposes to use these equations because, at the least, they are statistically reliable forecasting tools. 53 MR. GARDINER: Yes, that is correct. 54 MR. JANIGAN: As I understand it, these equations are a form of econometric model -- I hope I don't have to use that word too often -- and the econometric model as I understand it, is an equation that explains the outcome of a dependant variable by looking at the behaviour of independent variables. Have I got that right? 55 MR. GARDINER: That's right. 56 MR. JANIGAN: Okay. Now, in effect, there should be some causal relationship between the dependent variable and, in this case it would be the NAC, to the explanatory variables. 57 MR. GARDINER: Yes, that's correct. 58 MR. JANIGAN: Now, I'd like to look at the equations you've set out in C.1, tab 1, page 5, bottom of the page. 59 MR. GARDINER: Could I have the reference again. C.1, tab 1? 60 MR. JANIGAN: Page 5, bottom of the white pages. 61 MR. GARDINER: Of the white page. Yes, I have it. 62 MR. JANIGAN: Okay. And I assume equation number 1 is what Union terms the volume equation; am I correct on that? 63 MR. GARDINER: No. That is the use-per-customer equation. 64 MR. JANIGAN: The first equation is the use-per-customer equation? 65 MR. GARDINER: Yes. 66 MR. JANIGAN: And the second equation is the volume equation? 67 MR. GARDINER: Correct. 68 MR. JANIGAN: Okay. And in the first equation, I assume that what generated is a function of weather efficiency and price and price as stated in your evidence. 69 MR. GARDINER: Correct. 70 MR. JANIGAN: And the second equation is a function of weather over customers -- should that be efficiency over customers? 71 MR. GARDINER: No, no. The actual volume equation would read: "Volume is a function of weather, customers, and price," and then you would have an identity that would say "NAC is equal to forecasted volumes divided by your customers." But in order to get this in a nice way, the divide by customers is really the second step. 72 MR. JANIGAN: I realize I'm embarking into an area, probably, that's not my forte, but doesn't customers over customers equal one? 73 MR. GARDINER: That is my point. Not that it equals one, but it's a total volume equation. The total volumes is a function of the weather, the total number of customers, and the price, and you obtain forecasted total volumes. Once you obtain that, in the second -- in an identity, basically, to say how I turn the total volumes into the NAC, I take that total volumes estimate and divide it by the number of customers and then I get my NAC estimate. 74 MR. JANIGAN: Okay. Now, in interrogatory number J.34.33, we asked you to provide the data series that you used for your regression analysis for your model. And I wonder if we could bring forward the materials for cross-examination that we sent earlier. I wonder if I could have that marked as an exhibit, Mr. Chairman. 75 MR. MORAN: Mr. Chair, Exhibit M.5.1 is materials for cross-examination for VECC. 76 MR. SOMMERVILLE: Mr. Moran. 77 EXHIBIT NO. M.5.1: MATERIALS FOR CROSS-EXAMINATION FILED BY VECC 78 MR. GARDINER: I have it. 79 MR. JANIGAN: And on the first two pages of this exhibit, we've set out the results from the Excel spreadsheet that you provided to us for those two equations; is that correct? 80 MR. GARDINER: That's correct. 81 MR. JANIGAN: And on the first result, on page 1, that is the result of equation number? 82 MR. GARDINER: These are the volume equations, so this is equation 2. 83 MR. JANIGAN: This is equation number 2, okay. And on page 2, that's the result of equation number 1. 84 MR. GARDINER: On page 2, yes, that's correct, those are the use-per-customer equations. 85 MR. JANIGAN: Now, looking at the results at the top of the page, you have the results of an R-squared test, which I understand gives the degree of correlation between the independent variables of the model and the dependent variables so derived; am I correct on that? 86 MR. GARDINER: That's right. 87 MR. JANIGAN: Okay. And the T statistics give the significance of the independent variable in arriving at the result. 88 MR. GARDINER: That is correct. 89 MR. JANIGAN: And correlation and significance are not one and the same concept, as I understand it. 90 MR. GARDINER: That's correct. 91 MR. JANIGAN: Okay. You could have a high degree of correlation without the independent variable being statistically significant. 92 MR. GARDINER: That is correct. 93 MR. JANIGAN: Okay. An example that helped me understand this, that if you were trying to predict the number of students, let's say, in grade 7 in a given student population and you constructed a model where the independent variable was ownership of a backpack, you would get a pretty high degree of correlation between the students and the backpack, but ownership of the backpack would not necessarily be a good predictor of whether or not a student was in grade 7. 94 MR. GARDINER: That is correct. 95 MR. JANIGAN: Okay. I'm glad I understood it. 96 Now, just moving on to these statistics, if we look at the statistical significance of the explanatory variable in number 1, on page 1 -- oh, first of all, would you agree with me, given the degrees of freedom in your equation, that in order for your explanatory variable to be statistically significant, your T statistics should be at least equal to 1.8 or greater? 97 MR. GARDINER: I disagree with you. 98 MR. JANIGAN: Okay. 99 MR. GARDINER: On a purely statistical point of view, the question is: What level of confidence do you want to place on the statistic? But this is also a forecasting tool. And econometricians who are forecasters, they are challenged to come up with an equation that explains consumption. 100 Mr. Janigan is absolutely correct on making sure that the variables in the equation explain consumption. And traditionally, utilities, when looking at consumption behaviour, have used weather, prices, and efficiency. Australian beer consumption would correlate fairly well with natural gas consumption, but you can't use that, okay? Retail prices are an important demand driver. At the 95 confidence interval, the retail gas price on this table isn't significant, but it is at lower levels of significance. It has a negative sign, so the relationship between gas consumption and prices is proper. When prices go up, consumption goes down. 101 The coefficients that's obtained from these equations, especially the usage equations, can give us an idea of the price elasticity, and then we can compare that price elasticity to other published documents that have elasticities for products like natural gas delivery and other products, which suggests that it's inelastic. And I've done that and I'm in the -- I'm into the very price inelastic range of .1. That's why I accepted this variable. Because, as you remember, a number of years ago gas prices went sky-high, and they've come down, and they've gone up again. And in the company, people want to know, Well, when the price goes, what happens to the consumption? That's why I have it in the equation. 102 MR. JANIGAN: Okay. Let me get back to something I modestly understand, is that as I understand it, there are tables where, if you calculate your degrees of freedom and you look up for your T statistics, it gives a figure which is the threshold by which your variable has statistical significance or not. 103 MR. GARDINER: Mr. Janigan, I agree with you. If you go to the tables and look at that number, some of these things will not be at the 95 confidence. However, as I just stated, you don't throw the baby out when you want to wash the baby. 104 MR. JANIGAN: Let me establish that if we were looking at -- in a pure statistical analysis, if you went to the table, the level for significance is about 1.8, given your degrees of freedom; am I correct on that? 105 MR. GARDINER: I'll accept your looking up in the table, but forecasting isn't pure econometrics. You have to put it into practices. That's what I've been trying to say. 106 And the other thing I want to say about the regressions, there's another statistic at the top, you've seen it before, MAPE, mean absolute percent error. That is the statistic that shows how well these equations forecast annual volumes over time. How well did this equation, between 1990 and 2002, forecast annual volumes? Now, for residential '01, I'm within .8 of 1 percent. If you look across, you'll see rate 01, rate M2 commercial, 1.4; commercial, 1 percent, 2 percent, 1.5 percent. 107 As I mentioned yesterday, I'm more challenged with the industrial group, 4 percent. That's how good this volume equation forecasts. 108 Turn the page, please. You'll see the usage equations, and look at that MAPE. Again, very low percentages. I have copies of the charts that were also contained in that CD in which I provided all the information and all the statistics and everything that I used. It's all there. And I would like to use those charts to further my point. 109 MR. PENNY: I should point out, Mr. Chair, the CD that Mr. Gardiner referred to is the answer to J.34.33 that Mr. Janigan took this material from. It was provided in electronic form when the undertaking was answered -- the interrogatory, I should say. 110 MR. SOMMERVILLE: Thank you, Mr. Penny. 111 MR. GARDINER: If everyone has the charts in front of them, the first chart I have is the residential M2 actual volume -- 112 MR. MORAN: Perhaps we can mark this as an exhibit before the witness continues. 113 MR. PENNY: It's part of J.34.33. It's already been produced in these proceedings. 114 MR. SOMMERVILLE: I think that's right, Mr. Penny. I don't think it actually does need to be -- 115 MR. MORAN: It doesn't indicate anywhere that it is part of J. Since it's in electronic form, we don't have the CD in front of us and it's not very accessible. It just might be for convenient, that's all I'm suggesting. 116 MR. SOMMERVILLE: Do you have a problem with that, Mr. Penny. 117 MR. PENNY: Not particularly, as long as it is well understood that this is already evidence -- if it makes it more convenient to refer to it, that's fine. I want to make the point that it's already evidence in these proceedings. 118 MR. SOMMERVILLE: Fair enough. Give it a number, Mr. Moran. 119 MR. MORAN: Exhibit M.5.2, a series of graphs which are excerpted from undertaking -- sorry, I forgot the number of the undertaking. 120 MR. PENNY: J.34.33. Interrogatory, sorry. 121 EXHIBIT NO. M.5.2: SERIES OF GRAPHS EXCERPTED FROM UNDERTAKING J.34.33 122 MR. GARDINER: Mr. Chairman, if I could walk you through these charts. What I've presented here are the residential charts for both the volume equation and the actual use equation, and the first two charts show for residential rate M2, those are the residential customers in the southwestern part of the franchise. And as you can see on the first chart, this is total volumes -- actual total volumes being predicted by the equation. The actuals are in blue and the prediction estimates are in pink. 123 As you can see, the model picks up the seasonality very well, and that's because of the weather variable, and picks up, because of customer growth that's in the volume equation, the increase -- if you look at this, there's a subtle growth over time in the volumes. 124 If you turn to the next page, and the chart there is residential rate M2, actual use versus predicted use, and again you can see how well the model, the equation, is predicting actual use per customer using actual weather, the efficiency variable, and the price variable. You can see the slight declining trend that's present. You see the strong seasonality. That's why I use econometrics to do this. 125 I wanted to give you this visual picture because a visual picture is much more interesting than a MAPE. But what the MAPE says is that, on an annual basis, I'm within those small percentages. And the purpose of the equation is to say, what is going to be the annual volume or the annual usage in 2004. 126 The other rate cases, I didn't photocopy the other service classes and rates because essentially the charts looked very well, and if people go to the CD that was provided, they're all there. 127 MR. JANIGAN: Are you finished? 128 MR. GARDINER: Yes. 129 MR. JANIGAN: Okay. Let me go back to the T statistics again. In particular, I want you to look at page 1 of our evidence. I think it indicates in a number of different areas that the T-statistic value is below the 1.8 that I suggested. 130 MR. GARDINER: Mr. Janigan, it's irrelevant, as I mentioned. 131 MR. JANIGAN: That's not my question. Mr. Gardiner, I've given you your chance to explain why you don't believe those are relevant, but I'm merely asking you to confirm that the T statistics in the individual charts that you've provided are below the 1.8 confidence level that we spoke of before. 132 MR. GARDINER: Oui, Monsieur Janigan, yes. 133 MR. JANIGAN: Okay. Rate 1, if you look on the page 1, the retail gas price is 1.35, under M2, it's 1.28. 134 MR. SOMMERVILLE: Are those negatives in front of those numbers, is that negative 1.35? 135 MR. GARDINER: Yes. 136 MR. JANIGAN: And the commercial, under the rate 1, it's negative .19? 137 MR. GARDINER: Yes. 138 MR. JANIGAN: And the rate M2, it's .16, negative? 139 MR. GARDINER: That's correct. 140 MR. JANIGAN: And across the board on industrial, we have the .11, and down below for the coefficient of the industrial efficiency variable of a .27 and a 1.43. 141 MR. GARDINER: Yes. And as long as the signs are of the proper direction in the relationship, they can be accepted in the forecasting equation. 142 MR. JANIGAN: Yes. But in a traditional statistical model, this is below the confidence level that you would find on a table. 143 MR. GARDINER: Yes. At 95 percent confidence, they don't. But at other confidence levels they would be accepted. 144 MR. JANIGAN: It's your accuracy test that you would principally rely on, then? 145 MR. GARDINER: Yes. 146 MR. JANIGAN: Now, on the next page, page 2, there are also variables that are below what would ordinarily be the traditional confidence level. 147 MR. GARDINER: That is true. 148 MR. JANIGAN: Okay. Now, as I understand it, you only used the one equation for the development of the forecast for industrial customers. 149 MR. GARDINER: That is correct. 150 MR. JANIGAN: Now, after you have the two equations forecasted, I understand you take a simple average of the two results. 151 MR. GARDINER: Yes. 152 MR. JANIGAN: And in Interrogatory J.34.32, VECC asked Union to provide the evidence that statisticians commonly average two separate equations as a tool for forecasting, and I understand there was no independent statistical evidence to support that methodology. 153 MR. GARDINER: I'm a practicing forecaster. I'm always trying to improve my forecasting tools. Actually, the combination of the two resulted about two years ago when we started looking at the industrial sector, and we found that the usage equation wasn't performing very well and we discovered the volume. So then we said, Well, let's try it on a residential. We found a very strong relationship. And it makes sense. If you add customers, you get more volume. And then what we found is that because they did come up with different estimates, if we brought them together, could we be closer when the actuals came in? And we found that to be the case. 154 So we said, This is interesting. I have two strong tools. Instead of choosing one, marry them. It's a pragmatic approach, and the results so far are indicating that that approach is a better way of predicting the actual normalized volume. 155 MR. JANIGAN: But in terms -- 156 MR. GARDINER: It's not classical, it's a little Gothic, I admit, but it's a great tool. 157 MR. JANIGAN: This is something that you've derived, it doesn't seem to have any presence in any sort of statistical literature, but you think it works. 158 MR. GARDINER: It does work. 159 MR. JANIGAN: Okay. Now, in terms of the data set that you've used, in Exhibit 1 -- Exhibit C.1, tab 1, page 5 in your evidence on line 6 and 7, it's noted that Union relied upon data going back to 1991 for its NAC equations. 160 MR. GARDINER: Actually, it's 1990. 161 MR. JANIGAN: I'm sorry. It's corrected, I believe, on the blue page to 1990 to 2002. 162 MR. GARDINER: Yes, that's correct. 163 MR. JANIGAN: Now, in looking at your response to J.34.32, it's difficult to understand the use of only 12 years of data, given the comments there. It's set out at 34.32C, that: 164 "A longer time series enables the regression to better explain the variation that is present in the input-dependent variable. Increasing the number of observations is important for the heating degree day variable which is specified as a monthly matrix where each heating season month is a separate independent input variable. The longer the time series, the more data observations there are for each month degree day variable." 165 Given that kind of conclusion, why is only 12 years used? 166 MR. GARDINER: Well, I used 12 years because that is a data set that I have available on a consistent service rate class basis. Had I had data for the different rate classes and service classes that Mr. Janigan mentioned earlier, I would have probably gone back. But in the Centra case, there were consolidations in the data and I didn't have the breakout between some of the services classes. 167 The key point in the answer here is as was mentioned: The weather variables in the equation are separate heating degree days for each month, and I use the word "matrix," and I don't want to make this too complicated, but instead of having one time series that just says heating degree days with 159 numbers, there are nine series. One is for January. So every January, there's a heating degree number, and then zeros for the others. February there's a number, zeros for the other months. With 12 years, I only have 12 data points for January and 12 data points for February. That's why I go long. I want to have as many data points for those months as I can get in order to get a better estimate of what the weather coefficient is for January. 168 The result of using this data matrix, and actually the data matrix was developed by my precursor -- or my precursor was using that at Union Gas, and it's a very good approach, and using the data matrix, how we can get weather coefficients for each month so that the impacts for each month is not the same. That's very important for months like October compared to January. 169 So that's the reason why I used 12 years. If I had more, I'd use more. But next year there will be 13 years. 170 MR. JANIGAN: Now, would you agree with me that over the last 12 years has generally been a warmer-than-normal trend in weather? 171 MR. GARDINER: That doesn't matter. I use the actual weather. 172 MR. JANIGAN: yes, but over -- the last 12-year data series that you use reflects a warming trend in the weather. 173 MR. GARDINER: Right. And the model -- the equation says, What is the consumption for the actual weather? 174 MR. JANIGAN: Yes. 175 MR. GARDINER: When it's warm, you get less consumption; when it's cold, you get more consumption. 176 MR. JANIGAN: But if your data series has 12 years of entries and those entries reflect a warmer-than-normal pattern -- 177 MR. GARDINER: It's irrelevant. 178 MR. JANIGAN: Pardon me? 179 MR. GARDINER: Regression analysis using actual data, which is what you do, the relationship is between actual consumption and actual weather. 180 MR. JANIGAN: Yes. 181 MR. GARDINER: Whatever trend it is in the weather, whether it's warm or getting colder, doesn't matter. It picks up the relationship between consumption and weather. 182 MR. JANIGAN: You mentioned earlier the importance of an equation of a forecasting tool, that it's able to forecast accurately into a time period. I assume that's an objective? 183 MR. GARDINER: Yes, that's an objective? 184 MR. JANIGAN: Okay. And obviously these forecast equations are built to produce a NAC for levels that will occur in the test year. 185 MR. GARDINER: That is correct. 186 MR. JANIGAN: All right. Now, in order to test a forecast predictability of any econometric equation, usually a forecaster produces an ex-post forecast in which you would rely upon this forecast equation. 187 MR. GARDINER: That was done. 188 MR. JANIGAN: And when VECC asked Union for the NAC forecast accuracy in an ex-post forecast period, you gave your answer in J.34.32, I and J, to the effect that the ex-post forecast, I don't know whether you did one or not, but that the ex-post forecast of the model would be inaccurate, or it didn't produce the right results. 189 MR. GARDINER: An ex-post analysis was prepared, and the issue I mentioned in the interrogatory, the fact that models -- and that's one of the reasons why the T stat on the gas price is not that good -- is that we did see dramatic change in the retail prices. 190 One of the limitations of econometrics is -- challenge, if you want, is that if there's a lot of variation or a real big structural change in one of the -- in the variables, the estimated coefficient is sensitive to that. And what we saw with retail prices were incredible changes, you know, doubling of prices, which was not in the historical time base. If you look at prices in '90 to '98, they vary 5 percent, 10 percent, but not a hundred. So that is why when we looked at the ex-post, we said yeah, I mean, the ex-post analysis is going to say, yeah, it's not going to forecast very good because of this issue with the price variable. 191 MR. JANIGAN: So you did produce this forecast but you don't believe it had -- it accurately tested your model. 192 MR. GARDINER: I think the forecast that we have, the models are very good. To do an ex-post analysis to say, okay, drop out '99, 2000, 2001, where we have significant price change, run your model on a period when prices were rather stable and then forecast, yeah, you're going to get an estimate that's off. 193 MR. JANIGAN: You did that, then. 194 MR. GARDINER: Yes, we did. 195 MR. JANIGAN: I wonder if you could produce that. 196 MR. GARDINER: I don't have it with me, but we did do -- 197 MR. JANIGAN: Sure. Could you undertake to produce that? 198 MR. GARDINER: Yes. 199 MR. JANIGAN: Okay. Could I have an undertaking for that. 200 MR. MORAN: Mr. Chair, this would be Undertaking N.5.1. What description would you apply to this document? 201 MR. JANIGAN: This is the ex-post forecast of the NAC model, Union NAC model. 202 UNDERTAKING NO. N.5.1: TO PROVIDE EX-POST FORECAST OF THE UNION NAC MODEL 203 MR. JANIGAN: And as I understand it, you take the nine years of data, if you have 12 years of data series, you take the nine years of data and try to predict what's going to happen in the three years -- 204 MR. GARDINER: That's correct. 205 MR. JANIGAN: Okay. Is it your understanding of Union's position in this proceeding that there will be less volatility in retail gas prices in the future than we've seen over the past three years? 206 MR. GARDINER: That's the reason why, in the price forecast, we kept the price at the March QRAM and the current rates. We recognize that there is price sensitivity. But we also know from our equation that it's very insensitive. It takes very large price increases to move the NAC. Subject to checking my notes, but it takes, like, a 14 percent change in price to get a 1 percent change in NAC, for the residential, even smaller for the commercial. To move NAC, say, 5 percent, you need a 15-percent price increase. 207 So we recognize that there's price volatility uncertainty. Our price elasticity is in the right range so I kept it in. Plus I'm satisfied that I have a good tool. 208 MR. JANIGAN: Let me get back to my point. I assume there is no understanding, you don't have a belief that NAC retail gas prices are going to return to the stability that we've seen in the '80s or '90, let's say. 209 MR. GARDINER: Well -- 210 MR. JANIGAN: Let me rephrase that. Do you have any reason to believe that prices will not be as volatile over the next period of time as they were over the past three years? 211 MR. GARDINER: The price forecasts I've seen vary. They're not as stable as they were in the early '90s. But that's a good thing because as I move my forecast forward and I get more price variability, hopefully, when I redo this, I'll capture that. One of the problems in the past was I didn't have enough volatility. 212 MR. ROGERS: Just as a point of clarification for the industrial market. I think we would say that the supply has tightened in this period of time, going forward, into 2004, and what we see that as raising the bottom of the gas price and being up at a higher level. Now, the volatility from there would be driven by other market factors, but the bottom of the gas pricing has come up so that it remains high. And that's my point in my testimony, that it would remain high. Whether there would be volatility on top of that base because of lack of supply is something that we'll see, going forward. 213 MR. JANIGAN: Okay. But I'm just going to return to Mr. Gardiner. We had difficulty in doing an ex-post forecast, some difficulty in accepting the results of an ex-post forecast because of the volatility in gas prices over the next three years. If gas prices are going to be volatile for the next three years after this, why is your model going to be a good predictor for that period of time? 214 MR. GARDINER: As we go into the future, I'll re-estimate the model. The volatility in the price is a good thing from an econometric exercise because with the demand-price relationship, as the price changes, we should see consumption falling. And with prices moving, and I am seeing consumption moving, I should be able to better capture that. 215 One of the issues with early regressions done in the mid-'90s was price sometimes wasn't significant and had the wrong sign. It wasn't being picked up. Because of some of the volatility that we've seen in 2000, 2001, 2002, I'm picking it up. Agreeably, it's at a weak level of confidence, but it's the right sign, it's the right relationship, and elasticities are in the right order of magnitude. 216 MR. JANIGAN: Well, when are we going to get along in the time series that we can have confidence an ex-post forecast will pick up and measure the accuracy of the model appropriately? 217 MR. GARDINER: That, I don't know. It will be sometime in the future, but I don't know. 218 MR. JANIGAN: It would seem to me, Mr. Gardiner, that the reliability of your forecast is dependent upon the reliability of Union's ability to forecast the input variable accurately. 219 MR. GARDINER: Yes. 220 MR. JANIGAN: And if the historical data series of 12 years that's explicit in your explanatory variables does not repeat in the forecast period, the NAC forecast produced is going to be incorrect. 221 MR. GARDINER: They will be -- I mean, it's a forecast, and that's my challenge, to try and get it within the bull's eye. The price variable that is most volatile -- well, there's two, there's weather, but that's why we use the normal and we normalize back to the normal, we take actual and normalize back to the normal. And as you can see from the T statistics on the weather coefficients, they're very strong. So price, yes, it will move. However, it's inelastic; it takes 50 percent changes in price to get, like, a 1 percent change in consumption. 222 So I'm not too concerned about that. Retail price is now at around 40 cents, about there, a 10 percent change is 4 cents per cubic meter. I'm not expecting 60 cent retail at the burner tip prices for 2004. That's what a 50-percent change would be. If we're talking about price volatility in the 10 percent range because of the commodity price component which is only a portion of the total burner tip price, you know, if it changes by 10 percent, the NAC will be off by 7/10ths of 1 percent in the residential market. Yeah, I'll be off. If my forecasting tool says plus or minus 2 percent, 7/10ths of a percent, I'm on the bull's eye. 223 MR. JANIGAN: But as I understand it, Mr. Gardiner, it was the changes in the prices over the last three years that made it impossible to use the nine years of data to forecast those three years. 224 MR. GARDINER: The key point is there is if you go back and stop in '99, your regression will be based on information that says the retail burner tip price is somewhere between 19 and 20-something cents, 22 cents, just moving a little bit up. And then it went, like, 22, 25, you know, up to 40, and now it's fluctuating in that, you know, 37-40 range, and it's that movement from 20 cents -- and sometimes even, if we go back and look at the series, I believe there's some 17 cents in there. But generally speaking, it was quite flat around that 20 cents, and then we saw a big run-up and it's come down and now we're at a higher level. And that's the commodity price, which is a portion. I don't see this as an issue. 225 MR. JANIGAN: Okay. Mr. Gardiner, I want to push ahead to your reasonableness test. Is it reasonableness or reasonability? I can't recall. 226 MR. GARDINER: I think it's -- I think reasonability. 227 MR. JANIGAN: Reasonability? 228 MR. GARDINER: Yes. 229 MR. JANIGAN: As I understand it, once the equation forecasts for the NACs are produced, that you do this reasonability test, and it's set out at Exhibit C.1, tab 1, pages 7 and 8 and there's a discussion of this. 230 MR. GARDINER: Yes. 231 MR. JANIGAN: "To ensure that the methodology is reasonably representing reality, the first year of the forecast... is compared to an outlook for that same year." 232 And that outlook is how many months? 233 MR. GARDINER: Three. 234 MR. JANIGAN: Three months. 235 "If the forecast value is within 1 standard deviation of the outlook, then the forecast for that year and subsequent years passes the reasonableness test. If the forecast is above the top of the range of one standard deviation, then it is reduced to the top of the range. If the forecast is below the bottom of the range of one standard deviation, then it is increased to the bottom of the range." 236 Now, it seems to me that when Union makes this adjustment to the forecasting model, it is really the outlook that is being used to forecast rather than the econometric model; wouldn't you agree? 237 MR. GARDINER: No, I disagree. 238 MR. JANIGAN: Why is that? 239 MR. GARDINER: In the cases where the NAC estimate, and the NAC forecast estimate, as I explained to Mr. Thompson yesterday, who actually walked me through this, and I appreciated him walking us all through this, there are, as we mention now, a marriage of two equations, estimates from two equations; then we apply the incremental NAC and the incremental DSM, NAC impacts to the econometrically arrived at NAC. And if that forecast NAC goes through those goalposts that I was talking about, the upper and lower range of the reasonability test, there's no adjustment. 240 The reason I do this is because I'm using the reasonability test as saying, Well, I've taken two econometric equations, I've added things to it, is it possible I've done something that is putting me outside of a range? 241 So with this strong stable relationship that I described yesterday, it's a last-minute litmus test to say, Okay, is it reasonable? There were only two rate and service classes that were adjusted and brought down to within the range. And if you look at the size of the adjustment compared to the final NAC, it's a small percentage change, it's within my, you know, 2 percent bull's eye. 242 But the key point here is, I do this because I'm adding, you know -- there's some econometrics in it, there's some marketing, there's some DSM in it, and then we have this final NAC. 243 MR. JANIGAN: Mr. Gardiner, I assume that in the circumstance where your forecast doesn't coincide or arrive at one standard deviation of the outlook, you would adjust so that it does conform to the outlook? Would you agree? 244 MR. GARDINER: Yeah. There were five rate and class services where the final NAC forecast estimate was within the reasonability test range, and there were no adjustments made to that. 245 MR. JANIGAN: Okay. But the outlook is the determining factor here. In other words, you arrive at a result that doesn't coincide with the outlook -- 246 MR. GARDINER: I don't use the trend estimate of the NAC reasonability test, which could be obtained, and the only problem with that is it just gives me the outlook for '03. It says I have an estimate that's going to go out the window, is it going to go out the window or if I throw it, am I going to hit the ceiling? That's what it is. The reasonability tells me how big the window is and where it is. 247 MR. JANIGAN: Okay. But if it doesn't hit the window or it goes out the window, I don't know in your analogy which one it is, then you go back to the outlook. 248 MR. GARDINER: No. I just -- the adjustment that comes from the reasonability test brings you back -- tells you how much you have to adjust, and then that adjustment is applied not only to 2-03 but also to 2-04. And remember, the combination of all those things are going from 2002, your last year, to 2003 and 2004. So it's just a final check to see if possibly your equation and your -- other adjustments are leading you off. I have another set of information of the stable trends that say is it -- am I in the ballpark. And the final adjustments are very small. 249 MR. JANIGAN: Now, I want to turn to your accuracy forecast and your tests for accuracy. I note in your blue-page update, Exhibit C.1, tab 1, page 3, that your evidence notes that the accuracy of the NAC forecast continues to be high. 250 "On a customer-weighted basis, the overall forecast is within 1.2 percent year-to-date when comparing forecast versus actual NAC." 251 Now, is that the same comparison that we talked about earlier as MAPE, or is that a different comparator? 252 MR. GARDINER: No. This one is -- the actual NAC that's obtained for the -- at this time when we responded, we had -- I'd have to check the reference. But I actually have figures for up to August so I'll talk about up to August. You have your actuals up to August and then you have your September through December estimates, so that gives you an outlook of eight and four, and you compare that to your forecast estimates, so you get your forecast error. And then that's calculated into a percentage; in some cases it's a positive, in some cases it's a negative. So you do that for each rate and service class. 253 Then you apply weights to each class, and the weights are based on the number of customers in each class. So, for example, the residential M2 group has a fairly high weighting because there's 800,000 customers out of 1.2 million. So those weights are then applied to each of those little errors, and you do the multiplication and add up the results and you get the weighted, consolidated error. And that's what that 1.2 is, I believe, that's in the interrogatory. 254 MR. JANIGAN: Mm-hm. 255 MR. GARDINER: And I have in my notes -- bear with me. At August, the residential M2, the forecast error was negative 32 cubic meters, and that means the forecast was too high by 1.2 percent. The residential 01 was 0.4 percent. The commercials, the three commercials, the errors were negative 1.6 for M2, .1 percent for 01, 1.7 percent for commercial 10. And the industrial M2s, they're actually -- the forecast is too high, it was negative 8.7 and negative 7.9 percent. And that consolidated -- I thought I had that with me -- is less than 2 percent. But you can see that the equations and the method that we've been talking about is performing very well. We're within that 2 percent range currently. 256 MR. JANIGAN: Mr. Gardiner, just before we get into this again, I wonder if you could give me the relationship between this number that we're talking about here in this blue-page note of 1.2 percent, the number that is listed as MAPE on pages 1 and 2 of Exhibit M.5.1, and the graphs that you distributed this morning, the colour graphs -- 257 MR. GARDINER: Yes. 258 MR. JANIGAN: -- the part of J.34.33. Are they part of the same accuracy test? 259 MR. GARDINER: No. 260 MR. JANIGAN: Okay. 261 MR. GARDINER: Okay. The weighted calculation I just described -- 262 MR. JANIGAN: Yes. 263 MR. GARDINER: -- is really an AG indicated mean percent error. 264 MR. JANIGAN: Okay. 265 MR. GARDINER: So those that I was doing positive and some were negatively, and when you weight those you end up with a combined, because in the end, I'm trying to get the total volumes. 266 MR. JANIGAN: Okay. 267 MR. GARDINER: If you did an absolute error -- 268 MR. JANIGAN: That's the 1.2 percent we're talking about. 269 MR. GARDINER: Yes. The 1.2 is the weighted mean percent error. 270 MR. JANIGAN: Okay. 271 MR. GARDINER: Okay. I've found it. August, year-to-date, weighted mean percent error, 1.1 percent, negative. 272 MR. JANIGAN: So that updates the figure of 1.2 percent in the blue pages. 273 MR. GARDINER: Correct. 274 MR. JANIGAN: Okay. Now, can you contrast that with the MAPE number we see and the graphs that you distributed this morning? 275 MR. GARDINER: Yes, you can. I don't have -- I don't wish to do the calculation on the number. But you know we had positives and negatives -- 276 MR. JANIGAN: Yeah. 277 MR. GARDINER: -- the MAPE would be around 1.4, 1.5, as I have, you know, 1.2.4, 1.6.1, 1.7, and those are the residential classes and the commercial classes which are 99 percent of the customer base; and then the industrials which are very challenging at 8.779, but they're only a percent. So just doing this in my head, the MAPE would be an average of those numbers and it would be somewhere between 1 and 1.5. It would be very small. 278 MR. JANIGAN: And I take it the graph is the illustration of the MAPE? 279 MR. GARDINER: If you took -- yes, in effect, if you take the differences between the actual and the predicted value and sum those for the year, okay, and do the comparison to the actual consumption, that gives you an error and a percentage error which you can take in absolute terms, and then you do that for every year and then you average that and you get your MAPE. 280 MR. JANIGAN: Now, in order to arrive at the weighted average number and I guess also your MAPE, did you put the actual input factors such as weather and retail gas prices through the equation to produce the result which was then compared to the actual result? 281 MR. GARDINER: The comparison of the forecast accuracy to date is simply comparing the actual weather normalized consumption with the forecast. You don't need to go and look at your demand drivers. You just look at what was my weather normalized actual consumption against my forecast. 282 MR. JANIGAN: Okay. But in order to get your -- the forecast number you're comparing to, are you inputting into that forecast the actual numbers for weather and retail gas prices? 283 MR. GARDINER: No. I have my actual consumption. For August, I have eight months, I have my actual consumption. 284 MR. JANIGAN: Absolutely. 285 MR. GARDINER: I weather normalize that consumption using the weather coefficients obtained from my equation to get my normalized usage, and then I just do a comparison of the two. 286 MR. JANIGAN: Okay. 287 MR. GARDINER: Mr. Janigan, that step isn't relevant that you're talking about. 288 MR. JANIGAN: Okay. But I understood that this is a comparison of your forecast model with actual volumes, the predicted volume based on your forecast model and the actual volume that comes about. 289 MR. GARDINER: When -- 290 MR. JANIGAN: Let me stop you there so I can understand it. For the purpose of, let's say, 34.33, the graph that you have -- 291 MR. GARDINER: Yes. 292 MR. JANIGAN: -- when you're plugging in your figure for predicted volume, okay, from your models, from your equations -- 293 MR. GARDINER: Yes. 294 MR. JANIGAN: -- are you inputting the actual numbers for that particular period of time? 295 MR. GARDINER: I think I'm with you now, Mr. Janigan. In doing the estimation of the forecasting equations on a historical basis, and those are the charts that we were looking at, the actual prices, the actual weather, the efficiency index, which is a construct but there's an actual estimate for what the efficiency is in that month, so on a historic basis, all these things are actual and all the variables are in there. 296 The comparison of the forecast performances -- I've made my forecast for 2003 and 2004, I have now a set of NAC values. As I go through the year, the actual consumption is coming in, I only normalize for weather. I don't normalize for price, it would make the comparison look better, and I can't normalize for efficiency because I don't have a new data set for efficiency. So I can't change that. 297 MR. JANIGAN: Mr. Gardiner, ordinarily when you do a forecast, you don't have the actual weather or the actual price to plug in. You've got to estimate that. 298 MR. GARDINER: Yes. The forecast NACs are the weather-normal, the 2004 declining trend weather-normal; the efficiency index, which is a linear series of numbers, and that's projected into time, it moves very, very gradually, it's numbers like .7239 going to .7249. It moves -- like, the fourth decimal point. And then the price variable, as I mentioned yesterday, we basically kept the prices at the current levels, recognizing that there would be changes; but because of the pricing elasticity and the fact that we didn't expect very large prices, we said, Well, let's keep that constant. 299 MR. JANIGAN: But all your accuracy test seems to illustrate to me, Mr. Gardiner, is that your test is accurate if we plug in the very inputs that we are attempting to estimate. 300 MR. GARDINER: I correct for the known weather; I'm not correcting for the price. If I did, the price -- the accuracy would even improve. I'm at 1.1 percent. 301 MR. JANIGAN: Okay. Let me move on. 302 I have a couple of questions on Exhibit C.1, tab 1, page 10. 303 MR. GARDINER: C.1, tab 1, page 10. 304 MR. JANIGAN: Actually, I'm looking over on figure 2, on the next page. 305 MR. GARDINER: Okay. 306 MR. JANIGAN: Am I to understand that for each year in the normalized trend line there is a different number of HDDs that that is an input. 307 MR. GARDINER: The HDDs behind each of these NAC levels are the trend line estimates of the 2-04 declining trend. 308 MR. JANIGAN: I wonder if it's possible for you to provide a NAC trend for the historical period shown in figure 2, figure 3, and figure 4 and 5, by taking the actual normalized -- by normalizing the actual NAC by the constant forecasted HDDs for 2004? 309 MR. GARDINER: Yes, that can be done. 310 MR. JANIGAN: And to do it, generate it from the 20-year trend methodology and from the 30-year average. Is that possible? 311 MR. GARDINER: Yes, it's possible. 312 MR. JANIGAN: Okay. Could I have an undertaking on that, please. 313 MR. MORAN: Mr. Chair, that would become Undertaking N.5.2. I wonder if Mr. Janigan could just restate that. 314 MR. JANIGAN: I was hoping you weren't going to say that. 315 It would be an undertaking to provide the NAC trend for the historical period shown on figure 2, figure 3, figure 4, and figure 5, by normalizing the actual NAC by the constant forecasted HDDs for 2004, generated from the 20-year trend methodology and the 30-year average. 316 MR. MORAN: And those figures being in Exhibit C.1, tab 1, at pages 11 of 15 and following. 317 MR. PENNY: Just before we sign off on that, Mr. Chairman, it sounds like a fair bit of work. It intuitively evades me what relevance it would be to use 2004 HDD days, but it may be because I'm not following. But could Mr. Janigan enlighten us as to why this might be, in any way relevant. 318 MR. JANIGAN: Oh, boy, I was hoping you weren't going to ask that. 319 MR. PENNY: I caught one. 320 MR. JANIGAN: As I understand it, it's relevant to looking at the separation of the trend line based upon the temperature trend and based upon other factors that influence HDD. So you take the normalized 2004 figure, apply it across the 20-year trend and the 30-year average, and you arrive at the particular figures for those four graphs. So it's isolating it to one number, which, as I understand it, is the traditional method that was used in forecasting. Do I understand that, Mr. Gardiner? 321 MR. GARDINER: It was used because we had a 30-year average which didn't change. 322 MR. JANIGAN: Okay. 323 MR. GARDINER: What you have on the figures here, on page 11 of 15, figure 2, shows the NAC according to the 20-year declining trend, which is our normal. Our normal is the trend line. 324 MR. JANIGAN: Is that based on forecasted HDDs for 2004? 325 MR. GARDINER: Correct. This is the white-page version. 326 MR. JANIGAN: Yes. 327 MR. GARDINER: But accepting that, that was the HDDs, according to the trend line, for 2003 and 2004; and similarly, according to the trend line, the HDDs for 1993 and so on. 328 MR. JANIGAN: Okay. Would there be any difficulty in showing that in one undertaking? 329 MR. GARDINER: Yeah, it can be done. 330 MR. JANIGAN: Okay. 331 UNDERTAKING NO. N.5.2: TO PROVIDE THE NAC TREND FOR THE HISTORICAL PERIOD SHOWN ON FIGURE 2, FIGURE 3, FIGURE 4, AND FIGURE 5, IN EXHIBIT C.1, TAB 1, AT PAGES 11 OF 15 AND FOLLOWING, BY NORMALIZING THE ACTUAL NAC BY THE CONSTANT FORECASTED HDDS FOR 2004, GENERATED FROM THE 20-YEAR TREND METHODOLOGY AND THE 30-YEAR AVERAGE 332 MR. JANIGAN: Mr. Chairman, I'm moving on to another area which is -- 333 MR. PENNY: Sorry, I'm not sure where we got on that. The fact that it can be done isn't an explanation for why it's relevant to be done, and I still haven't heard any explanation of why we ought to do this. How is it going to help? 334 MR. JANIGAN: Well, it helps in terms of isolating the amount of -- the trend that's produced from the temperature, the different techniques of dealing with temperature versus the other factors that are associated with arriving at the HDDs. 335 MR. PENNY: Well, if what Mr. Janigan wants is to have these figures done as if we were still using 30-year average, that would be one thing, but that, as I understand it, is not what he's asking for because he wants us to use the 2004 HDDs for everything and that, of course, isn't the way that it would be done. 336 MR. JANIGAN: No. I mean comparing the 20-year trend with the 30-year average. 337 MR. SOMMERVILLE: And isolating the weather normalization outcome from all other inputs to the HDD determination. 338 MR. JANIGAN: That's correct. 339 MR. PENNY: All right. Just so we're clear, then, what we're going to do is generate these figures, isolating only the variable of weather normalization. 340 MR. JANIGAN: I believe that's correct. 341 MR. SOMMERVILLE: That's my understanding. 342 MR. PENNY: All right. 343 MR. GARDINER: Would tables do instead of figures? 344 MR. JANIGAN: I think that's fine. 345 MR. SOMMERVILLE: I think that's on the record. Mr. Moran, we won't tax you with repeating that. Is that satisfactory to the parties? 346 MR. PENNY: Yes, thank you. 347 MR. SOMMERVILLE: Did you give it a number? 348 MR. MORAN: N.5.2. 349 MR. SOMMERVILLE: Mr. Janigan, you're indicating you're moving into a new area. 350 MR. JANIGAN: Yes. 351 MR. SOMMERVILLE: We'll take our break now and resume at 10 minutes after 11:00. 352 MR. PENNY: Just before we go, Mr. Chairman, Mr. Janigan had given us a one-hour estimate, and I know you indicated that there was an expectation about sticking to those. Could Mr. Janigan give us an indication of how much longer -- he's now been an hour and ten minutes -- how much longer he's going to be? 353 MR. JANIGAN: Mr. Chairman, I'm afraid the fault is all my own. My consultant gave me an estimate of one to two hours on the first day, and I was wildly optimistic in putting one hour in. I would expect that I will be another three-quarters of an hour to an hour. 354 MR. SOMMERVILLE: Okay. That does take us outside of the -- 355 MR. JANIGAN: It does. 356 MR. SOMMERVILLE: -- where we thought we were. Optimistic estimates, we're all old enough to know that they're not worth very much. So if we can try to be a little more pessimistic about them, that would be helpful. 357 MR. JANIGAN: I will do that. 358 MR. SOMMERVILLE: And we'll carry on bravely. We are determined to complete this panel today, and that will require cooperation, as I've indicated, from all parties to try to, how shall I say, meet that target, as I said, all parties, including the panel. We accept your explanation, Mr. Janigan, and we will live with it. We'll resume at 10 minutes after 11:00. Thank you. 359 MR. JANIGAN: Thanks, Mr. Chair. 360 --- Recess taken at 10:50 a.m. 361 --- On resuming at 11:15 a.m. 362 MR. SOMMERVILLE: Please be seated. Thank you. 363 Mr. Janigan. 364 MR. JANIGAN: Thank you, Mr. Chairman. 365 I'd like to move on to the issue of transactional services, and I believe that's Mr. Poredos' area of speciality. I'd like to deal first with the magnitude of the transactional services -- revenue that's projected. 366 As I understand it, your evidence at C.1, tab 3, page 6, in the white pages -- 367 MR. POREDOS: The white-page evidence? 368 MR. JANIGAN: The white-page evidence, C.1, tab 3, page 6. It's indicated there that S&T transactional -- the S&T transactional services market has dramatically declined over the last few years; is that correct? 369 MR. POREDOS: Yes. 370 MR. JANIGAN: And on the next page several reasons are advanced for this dramatic decline, and these include Enron, the reduction of the number of credit-worthy counterparties, reduced summer/winter price differentials for natural gas, and high natural gas commodity prices. 371 MR. POREDOS: Yes, that's correct. 372 MR. JANIGAN: Okay. And these seem to be factors that should be affecting the S&T business across the board, regardless of company. 373 MR. POREDOS: That may very well be so. 374 MR. JANIGAN: I would like to deal with the revenue portion of Union's S&T assets. To do so, if you could turn up Exhibit A, tab 15, schedule 2, page 4, which is the Standard & Poors report that was filed by Union in this proceeding. 375 MR. SOMMERVILLE: What tab was that, Mr. Janigan? 376 MR. JANIGAN: Tab A -- Exhibit A, tab 15, sorry, schedule 2, page 4. 377 MR. POREDOS: Within the exhibit, page 4, sir? 378 MR. JANIGAN: Yes, page 4 within the exhibit. 379 MR. POREDOS: Yes, I have that. 380 MR. JANIGAN: And it indicates in the Standard & Poors report that the transmission, at the top of the report: 381 "The transmission and storage system are valuable assets and an integral part of Union's operations. The transmission and storage system interconnect with six major pipelines: TransCanada Pipelines, Great Lakes Gas Transmission, Panhandle Eastern Pipeline, Vector Pipeline, MichCon, and Consumers Energy. The combination of six major interconnections and access to large storage facilities (Enbridge Gas Distribution also has storage facilities near Dawn with 90 bcf/d of capacity), have resulted in Dawn becoming a major North American storage hub. Dawn has about 150 bcf of working capacity and can deliver over 2 bcf/d for storage customers." 382 It goes on: 383 "The company transports over 1.2 trillion cubic feet of gas annually. 384 "However, like other hubs, the ratio of market activity to gas in storage has grown significantly. Union Gas provides storage, balancing, gas loans, and peaking services to more than 100 customers and, in 2001, handled an average of 11.9 bcf/d of title transfers. Union is rated very highly by its customers for customer service and marketing effectiveness. Union Gas utilizes about 56% of the storage capacity for its own distribution operations, which helps to offset the effects of warm weather on the company's earnings." 385 It seems from that statement that Standard & Poors notes that Union operates storage and transactional services in an integrated market as Union is connected with six pipelines; is that correct? 386 MR. POREDOS: Yes, that's correct. 387 MR. JANIGAN: And one of the entities that is cited in this quote from Standard & Poors is Enbridge Consumers Gas; correct? 388 MR. POREDOS: Yes. 389 MR. JANIGAN: Does Union have knowledge of a dramatic decline in Enbridge's S&T revenues? 390 MR. POREDOS: It would be speculative of me to try and talk to the revenues of Enbridge, their transactional revenues. However, we did realize that this was an issue that was going to come up. We did try to contact Enbridge to see if we can get some information on this. We did get some basic information but did not get any detail, which, frankly, is not surprising in a competitive market. I wouldn't be discussing strategy, marketing position, pricing, and so forth or value of assets with a direct competitor. 391 MR. JANIGAN: Okay. Now, according to the evidence -- I wonder if you could turn up my Exhibit 5.1. 392 MR. POREDOS: Is that the handout? 393 MR. JANIGAN: Yes, it is, on page 3. 394 MR. POREDOS: Yes. 395 MR. JANIGAN: This is from the Enbridge proceeding, RP-2001-0032. They forecasted the S&T for 2002 of about 8.4 million, which was subject to the 90 percent sharing; is that correct? 396 MR. POREDOS: Yes. That's what I see on this page, yes. 397 MR. JANIGAN: Okay. And in the Enbridge 2003 rates case, Enbridge made an ADR agreement for 2003 which also addressed S&T revenues, and in the context of that agreement, Enbridge agreed to lock in $8 million for their S&T revenue requirement? 398 MR. POREDOS: That is my understanding from what I've read, yes. 399 MR. JANIGAN: In the 2004 proceeding, effectively, there was no S&T forecast, but do you understand that Enbridge, for 2004, continues to guarantee an $8 million reduction in rates for the S&T revenues? 400 MR. POREDOS: No, I do not have any knowledge of that. 401 MR. JANIGAN: Okay. And do you think it's odd that Enbridge, at a time when you're projecting this dramatic decline in revenue, that Enbridge is not reflecting a dramatic decline in the amounts that it is embedding for S&T revenues? 402 MR. POREDOS: Mr. Janigan, I think that if we look at the page 3 that you've handed to me, the updated evidence there was the end of 2001, and if you look at our actuals in 2002, which this reflects or forecasts forward, in fact, 2002 was an excellent year. And I'll agree to that. 403 The issue for Union is that we see the market, and I can't speak to Enbridge's view of the market, but Union does see the market as contracting. There are fewer players, as I mentioned yesterday. The traditional counterparties that we have are much fewer. Their credit ratings are a lot lower. And when credit ratings are a lot lower, you move your capital to where you make the most return, and in a lot of cases, those have moved capital out of the trading business; in fact, a lot of companies have gotten out of the trading business, including Aquila, Morant, Enron obviously, PG&E, and others. 404 As I mentioned, I think, previously in my evidence, with some of the major companies, including Aquila, including CMS Energy, Calpine and others, we did about $32 million of business a year, and this year we will do $8 million. That is a dramatic difference for us. 405 I think the other issues in the markets are there are lower demands. The northeast is well overbuilt on electricity, and a lot of those companies that have purchased those assets, capacity to feed the electric generation, physical gas to feed electric generation, will not be doing has in 2003 and 2004. That puts excess assets into the marketplace which now becomes more competitive. 406 A great example of the competitive market is we had originally forecast in white page that Enbridge would be putting in for about 170,000, 10(3)s, or sorry, gJs, I believe, of capacity on our Dawn-Trafalgar system. To our understanding, they went and signed for a third party for that same amount. It's a competitive business. We see the market going down. 407 The other issue from the S&T standpoint is we've also reduced our assets that are available. There's been six pJs of physical assets that have been moved off, three of it going back into the in-franchise market, three of it because we're not going to sign long-term contracts on some existing storage capacity that we inherited when we were merged with Centra. 408 MR. JANIGAN: But Enbridge operates in that same market, does it not? 409 MR. POREDOS: Yes, they do. 410 MR. JANIGAN: And they're subject to the same factors that we listened to earlier that cut across the industry? 411 MR. POREDOS: They may be. It's the same market. But, again, they may have a different type of customer. They may be going after a different type of customer. They may be operating in a different way. They may have had an advantage over the last year or two that we haven't. I can't speak to those. 412 MR. JANIGAN: Does it concern you that they are projecting and locking in amounts that do not reflect a dramatic decline in S&T revenues? 413 MR. POREDOS: I'm sorry? 414 MR. JANIGAN: Does it concern you they are locking into their rates numbers that don't reflect declines in S&T revenues? 415 MR. POREDOS: I'm not sure it concerns me, but I may be concerned if I was Enbridge perhaps. I can't speak to that. 416 MR. JANIGAN: Okay. Now, according to Standard & Poor's comment, Union is providing services to markets in the United States, and you're likely competing with S&T providers in the States, such as Michigan Consolidated. Do you monitor your competitors to see how they're doing in marketing their S&T services? 417 MR. POREDOS: Yes. We have market intelligence. 418 MR. JANIGAN: Is Michigan Consolidated reporting a decline in transactional service revenues? 419 MR. POREDOS: I don't have that specific information, sorry. 420 MR. JANIGAN: Is that something you are able to find out? 421 MR. POREDOS: I suspect if it's in a public document, we can probably find it. 422 MR. JANIGAN: I hesitate to ask you to undertake this. Does this involve a large search to do so? 423 MR. PENNY: What I think Mr. Poredos is saying is that he doesn't personally have that information, and I don't even know whether this is true, if Michigan consolidated is a public company, it may be that there's something in their annual report; it may not be. I don't know. But if that's so, in any event, Mr. Janigan, in my submission, can go and find the MichCon annual report just as easily as we can. 424 MR. JANIGAN: I'll move on. 425 I wonder if I can take you, once again, to the Standard & Poor's report, where they set out some comment concerning the competitive position, and that's, once again, at Exhibit A, tab 15, schedule 2. The second part of that paragraph: 426 "Transmission system and storage assets are quite strategic as the Dawn hub has become an integral part of the overall delivery system of natural gas from the West to markets in the Midwest and Northeast. The system's key competitive advantage is its large storage capacity, which enables the pipeline interconnections to be used efficiently. The fact that Union's facilities can efficiently serve the Midwest and Northeast U.S. is attractive to customers (and Duke) who are active in both areas. The competitive advantage of Union's storage and transmission assets is the combination of location, market liquidity, operating flexibility, and access to high-volume endusers. Thanks to Union's established position, these factors would not be replicated easily by competitors." 427 The statement seems to imply that Union has a strong competitive advantage in relation to other competitors. 428 MR. POREDOS: That may very well be true, Mr. Janigan, in the long run. What my evidence suggests is that in the test year, the opportunities will not be there to generate the amount of revenue that we've generated in the past few years. 429 MR. JANIGAN: Well, your evidence, I believe, on the white pages, says that there's been a dramatic decline over the last few years, and this Standard & Poor's report doesn't seem to be aware of any dramatic decline in revenue for Union. 430 MR. POREDOS: If I understand correctly, they were using data from 2001, in the paragraph above, Mr. Janigan, so I'm not sure whether they've got up-to-date data or not. I can't speak to how they've analyzed the market. 431 MR. JANIGAN: This report was issued in May 2003, I believe. 432 MR. POREDOS: I agree. But in the -- 433 MR. PENNY: No, it was not. 434 MR. JANIGAN: Sorry. It was issued on February the 12th, 2003. 435 MR. POREDOS: What I was referring to in the second paragraph, Mr. Janigan, I think it's the second sentence that suggests that: 436 "Union Gas provides storage, balancing, gas loans, and peaking services to more than a hundred customers, and in 2001, handled..." 437 That's the reference. 438 MR. JANIGAN: Yes. 439 MR. POREDOS: So if we're speaking about 2001 going into 2002, I agree. 440 MR. JANIGAN: Yes. 441 MR. POREDOS: The opportunities were there. In fact, we captured a large amount of those opportunities and created, if I remember the number correctly, and it's in my evidence, somewhere around $17 million of transactional revenue during that period. What we're saying today is that market is not available to us. We're saying that the market has changed dramatically enough that we will not be able to generate those values. However, if we were able to, and the market changed, those positive values would go through the deferral accounts and be shared with customers 25/75. So the customers aren't at risk. 442 MR. JANIGAN: So the window of this dramatic decline, then, is not necessarily the last few years, it could be shorter than that; is that what you're saying? 443 MR. POREDOS: The whole market has changed since the Enron failure that's caused a whole bunch of other items in the marketplace to take place. All energy companies are being affected by credit limits. 444 MR. JANIGAN: But surely, that is something that Standard & Poor's would be aware of when they're issuing a report in February of 2003; would you not agree? 445 MR. POREDOS: I would suspect they are, and I'm pleased to hear they have such a high regard for our company. 446 MR. JANIGAN: And they've made no mention of that in terms of impact on S&T revenues? 447 MR. POREDOS: Mention of? 448 MR. JANIGAN: The Enron fiasco. 449 MR. PENNY: Sorry. With great respect, this is in a section of the report that talks about competitive position. It's got nothing to do with revenues or forecast revenues. 450 MR. JANIGAN: Yes. Is there anything in the Standard & Poor's report that, in fact, takes the Enron debacle and uses it to impact on S&T revenues? 451 MR. POREDOS: I can't speak to that because I haven't necessarily read that particular report. 452 MR. JANIGAN: Okay. Let me move on. Has Union made any changes to the contracts associated with S&T services that may be generating a decline in S&T revenue levels? 453 MR. POREDOS: Some of the services may have changed. That does not necessarily provide that the resultant would be lower revenues. 454 MR. JANIGAN: Have you put any constraints, for example, on market hub contracts? 455 MR. POREDOS: Yes. Union has limited the amount that people can carry in a market hub, because people were using the market hubs as a cheap storage over winter and that wasn't available to them last year. So we wanted to make sure the customers understood that that market hub service was never meant to be used in that way. 456 MR. JANIGAN: What's the impact on revenue of that -- 457 MR. POREDOS: I believe that change is probably estimated at about a million dollars. I think it's in my evidence, in fact. 458 MR. JANIGAN: Now, on Exhibit C.1, tab 3, page 7, I believe, of the white pages -- 459 MR. POREDOS: White page or blue page, sir? 460 MR. JANIGAN: I believe that's white pages. 461 MR. POREDOS: Okay. Sorry, tab 3, page 7? 462 MR. JANIGAN: Yes, where you talk about title transfers. 463 MR. POREDOS: Yes. 464 MR. JANIGAN: And you indicate that title transfers are down, and you reference a 60 percent decline by looking at the last quarter of 2001 and the first quarter of 2003. 465 MR. POREDOS: Yes, sir. 466 MR. JANIGAN: Okay. Now, the first quarter of 2001, which is January, February, March of 2001, was this a period that was warmer than normal? 467 MR. POREDOS: Yes, I believe it was. 468 MR. JANIGAN: And the third quarter of 2003, which is October, November, December of 2003, was that period colder than normal? 469 MR. POREDOS: 2003, sir? 470 MR. JANIGAN: Yes. 471 MR. POREDOS: We're not there yet. 472 MR. JANIGAN: The third quarter would be -- 473 MR. POREDOS: 2002? This summer? 474 MR. JANIGAN: Let me see here. Let me compare -- I think I've -- I guess let me just ask the question directly. 475 Why don't you compare the same quarters in addressing the liquidity decline? 476 MR. POREDOS: Why didn't I? 477 MR. JANIGAN: Yes. 478 MR. POREDOS: It was a matter of the time when we were writing the evidence, I just chose a period where we had high transactional activity at Dawn to show the reference to another period during 2003 which was more up to date, and during the period I was involved in the evidence. That information is posted regularly on our web site, so it's information the public has. Customers can see that information. 479 MR. JANIGAN: I'm just wondering if your choice of quarters also influences, in a significant fashion, the evidence of the decline. 480 MR. POREDOS: I don't believe it does necessarily. But you would have some greater activity during the October period because of storage balancing, and perhaps during the early winter or into the spring because of people, again, trying to get gas in if they're short or long before they go into the summer. But normally the growth is sort of general and standard across the piece. It's been growing for a lot of years up until that point. And at that point the activity went down quite dramatically, and continued to be seen on our web site. 481 MR. JANIGAN: Now, there's been a transfer of the S&T service function from Union's Chatham office to Duke Energy in Houston, Texas; am I correct on that? 482 MR. POREDOS: The sales function has moved, yes. 483 MR. JANIGAN: Okay. And what was the date of that transfer? 484 MR. POREDOS: I believe the official date was this February 2003, subject to check. 485 MR. JANIGAN: What was the date again? I'm sorry. 486 MR. POREDOS: February 2003. 487 MR. JANIGAN: And are the Texans doing as good a job as their Canadian predecessors? 488 MR. POREDOS: I don't think I can answer on a comparison of Texans and Canadians. I can tell you that the S&T salespeople in the Houston office are doing just as good a job as our people in Chatham. Our customers are North America wide, and quite frankly, there are fewer customers today. 489 What's happened is regardless of where the salesman sits, they have to go visit the customers in Toronto, in Quebec, in Houston in the northeast, and in Calgary. So the place of where those salespeople sit really isn't that important. In fact, not to your surprise, there probably isn't many customers in Chatham. 490 The issue for us is that with Duke consolidating all their sales staff in one area, it created a functional approach to the sales. It coordinated it, centralized it in one area, and it was easier to take care from a sales function, a management function. It had nothing to do with the way we sell the services we provide or the assets we provide. 491 MR. JANIGAN: Is it possible to do a comparison of the revenues, S&T revenues, before and after the transfer? 492 MR. POREDOS: I'm not sure what that would prove, Mr. Janigan. The market is the thing that's driving this, not the salespeople that are in the seats. 493 MR. JANIGAN: One of the things we like to look at is if this -- I think, as you indicated earlier, some of the factors that are influencing S&T revenues are present both when you had the S&T being handled in Chatham and when you had S&T handled in Houston. 494 MR. POREDOS: I didn't say that. I thought I said it was the market conditions, not the salespeople. Did I misunderstand what you were just saying? 495 MR. JANIGAN: no, no, but those market conditions have been in effect both when you had S&T handled in Chatham and S&T handled in Houston. 496 MR. PENNY: The reporter is struggling here, I think it's because you're both speaking too quickly. 497 MR. SOMMERVILLE: Thank you, Mr. Penny. 498 MR. POREDOS: Could you repeat the question, please. 499 MR. JANIGAN: Okay. Is it possible for you to do a comparison between the S&T revenues prior to the transfer to Houston, and the S&T revenues following the transfer to Houston? 500 MR. POREDOS: We can provide the data. I mean, that's information that's already submitted within my evidence, and will be updated -- for 2003 as we submit the final year's numbers. I just don't know what the relevance is from a standpoint of how you would segregate or I would segregate the impacts of the marketplace from the impacts of the salespeople or where those salespeople are sitting. 501 MR. JANIGAN: That's a matter for interpretation. But just in terms of the raw data, you're able to produce revenues following the transfer and revenues prior to the transfer, in an appropriate period? 502 MR. POREDOS: Again, it will be very difficult to do that from a standpoint that some of our contracts that were signed back in 2002 for storage will run across a physical period, April to April. So are we saying that that now is an issue that is caused by salespeople? I don't think you're going to get an appropriate comparison here, and that's why I'm a little bit concerned that it will be an arbitrary comparison because you cannot distinguish between the market and the salespeople necessarily, especially since we've only had about six months of salespeople sitting in Houston. I don't think it's appropriate. 503 MR. JANIGAN: That's my principal concern is the amount of months of data may not be sufficient. Let me go at it from a qualitative fashion. 504 How many employees were working in this function before the transfer? 505 MR. POREDOS: There were four direct salespeople and three support people at the time. 506 MR. JANIGAN: And were they working exclusively for Union? 507 MR. POREDOS: They were working exclusively for Union. 508 MR. JANIGAN: And in Texas, how many people are there and how many are working exclusively for Union? 509 MR. POREDOS: There are four people in Texas, and they are virtually exclusively working for Union. 510 MR. JANIGAN: Okay. So they don't sell the other stuff of the Duke empire. 511 MR. POREDOS: They may sell some, but it's a smaller proportion in total. Their sales jobs have been reorganized, and some of the stuff they had on their plates before were shifted to others in Houston. 512 MR. JANIGAN: Now, do they have experience in -- well, first of all, are the employees the same employees? 513 MR. POREDOS: No, they're not the same employees. 514 MR. JANIGAN: Okay. Do they have the same years of experience in selling services for the Dawn hub? 515 MR. POREDOS: I can't speak directly to the number of years of service those employees have. I do know they've been with Duke for quite a long period. And they have similar experience to the salespeople that work for Union. 516 Duke Energy has transmission lines that are regulated under FERC. Most of their capacity has to be posted. They also have storage facilities that are regulated under FERC or one of the state regulations. They sell those services the same way as we would sell services. 517 The issues for Duke was you had two sets of salespeople talking to the same customer. We don't need that, and for efficiency's sake, we consolidated the sales force. The consolidation or the location of that sales force was decided to be in Houston so that they could have all the salespeople under one roof. 518 My understanding is that the Union salespeople were offered positions in Houston but they did not accept the transfer. 519 MR. JANIGAN: Now, those four persons, though, that are in Houston, they are selling pretty much exclusively for Union. 520 MR. POREDOS: Pretty much. 521 MR. JANIGAN: Okay. I wonder if this is a question for the O&M panel. I'd like, at some point in time, to get a breakdown of the cost that Union pays for S&T services in Duke in Houston versus the cost in Chatham. 522 MR. PENNY: You may recall, Mr. Janigan, if I may speak to that, Mr. Chairman, that there's a panel that is going to be dealing with affiliate relations, and that panel is aware of this function and the issues that you've raised associated with that. 523 MR. JANIGAN: Okay. Thank you, Mr. Penny. 524 I wonder if you could turn up Exhibit A, tab 12, page 31, the Union annual report. It's Exhibit A, tab 12, page 31. This deals, in number 15, with related party transactions. And we'll be having some questions for the affiliated panel on most of this area, but I want to just touch upon, briefly, some aspects of this that deal with S&T. It sets out there in 15a that: 525 "The Company purchases transportation services at prevailing market rates and under normal trade terms from commonly controlled companies. During the year ended December 31, 2002, these purchases totalled 74 million," versus 2001 of 53 million. "The Company also provides storage and transportation services to commonly controlled companies under normal trade terms. During this year, this revenue totalled $11 million," in 2001 it was 4 million. 526 The related parties are identified in J.1.23, the interrogatory -- I don't think you need to turn it up -- as Duke Energy Marketing Limited Partnership, Engage Energy Canada LP, Westcoast Energy (US) Inc., and Westcoast Energy, Inc. 527 Actually, I'm going to ask you to turn up that interrogatory, 23, now. I realize it's not your interrogatory. It's for a slightly different purpose that I ask the question. 528 MR. POREDOS: Yes, I have that, sir. 529 MR. JANIGAN: I wonder if I could direct you to the page that sets out transactions and prices and price range. 530 MR. POREDOS: Is that the second page you are directing me to? 531 MR. JANIGAN: That's correct. Do you know, or is this a question I have to ask of Ms. Elliott, is the price range set out in C the price range of the market or the price range of the transactions with the affiliates? 532 MR. POREDOS: I believe that that is the price range with the companies that we did the business with. 533 MR. JANIGAN: Okay. All right. I think I will have to have this line of questioning directed to the other panel because of that. 534 MR. POREDOS: Okay. Are we completed with this exhibit, then? 535 MR. JANIGAN: Yes, sorry. 536 Now, I want to deal with the topic of sharing of transactional services revenues, and I want to particularly deal with the sharing proposal that is advanced by Union in this proceeding, in particular, the proposal that the 2004 revenue requirement will be reduced only by the EBRO 499 S&T revenue forecast, and that any variance between the actual 2004 S&T revenue and the EBRO 499 forecast be disposed of on a 75/25 basis in favour of customers. And I believe you gave the answer yesterday to Mr. Warren in his examination that the reason for Union's position was essentially risk to the shareholder? 537 MR. POREDOS: That's correct. 538 MR. JANIGAN: Okay. And the last time Union generated rates based on cost of service at rate application was in EBRO 499? 539 MR. POREDOS: I believe that's correct. 540 MR. JANIGAN: And Union has produced a forecast for 2004 S&T revenues that was reviewed earlier in the proceeding? 541 MR. POREDOS: Yes. 542 MR. JANIGAN: Is it likely that this forecast made this year, 2003, is a better prediction of 2004 S&T revenues than a forecast made in 1998 for EBRO 499? 543 MR. POREDOS: If you're asking if the forecast we did today is a better reflection of what the proposed revenues would be in 2004 than the one we did in 1998, the answer is yes. 544 MR. JANIGAN: Okay. And if we were following the methodology in EBRO 499, then the 2004 forecast would be the input for the purposes of the sharing mechanism? 545 MR. POREDOS: When you speak on the methodology, could you just elaborate what you mean by the methodology? 546 MR. JANIGAN: Essentially, the forecast is embedded, and 90 percent of that embedded goes to the customers and then the excess would be shared on 75/25. 547 MR. POREDOS: Yes, that's correct. If you took the specific methodology that was done in 499, that would be the split. 548 However, having said that, it's a different market today, it's a different outlook. So to be embedding huge numbers into the rates, we believe that is not appropriate. We believe that we have reflected within our evidence and within our position the appropriate balance between customer and shareholder that will benefit all. 549 MR. JANIGAN: Okay. I just want to deal with Exhibit M.5.1 again, our handout. 550 MR. POREDOS: Yes. 551 MR. JANIGAN: I'm now on the topic of the long-term storage premium. We have an excerpt from the EBRO 499 Union Gas settlement agreement on page 8 of that exhibit. 552 MR. POREDOS: Yes, I have that. 553 MR. JANIGAN: That: [as read] 554 "All parties agree to the 1999 forecast of transactional services margin, excluding the 1990 forecast long-term storage premium of 2.571 million as discussed further below, of 5.561 million, and further agreed that the sharing of S&T transactional services margin should be consistent with the mechanism approved for the Board for Consumer Gas in EBRO 495, and specifically, this results in a sharing of the base forecast 90/10, while any variances in the forecast are shared on a preapproved basis 75/25 of ratepayers." 555 That was for the S&T services. On the next page it's indicated that: 556 "Union has forecast a market premium, 1999, of 2.571 million," and effectively this was to be credited to the ratepayers. 557 Would you agree with that? 558 MR. POREDOS: It's written in the decision or the agreement, yes. 559 MR. JANIGAN: Okay. 560 MR. POREDOS: I would also add, though, when you go forward, the decision was made, I believe, in 2001 by the intervenors and the Board that the long-term premium be shared 25/75 in favour of the customers. 561 MR. JANIGAN: We're going to get to that. 562 MR. POREDOS: Okay. 563 MR. JANIGAN: If we look at decision EBRO 486-02, which is set out also in the material at page 21. 564 MR. POREDOS: Yes, I have that. 565 MR. JANIGAN: That is dealing with this issue in paragraph 2.022, it's noted that Union argued, and this was dealing with, once again, the market premium question: 566 "Union argued that there were unlikely to be negative cost consequences to infranchise customers since in its proposal, they would receive the benefits from the economic rents arising from the difference between the market price --" 567 MR. SOMMERVILLE: Mr. Janigan, slowly, please. 568 MR. JANIGAN: Sorry. 569 ... market price and the cost-base rates for storage. Union submitted that this allocation of the rents was appropriate since infranchise customers had paid for the development of the storage." 570 Now, is it not still the case that customers still have paid for the development of those assets? 571 MR. POREDOS: I agree with the statement. It's in the decision. From my standpoint, though, we agree with the position that was taken at that point in time. We agree that the sharing that was taken -- was agreed to at that point in time of 25/75 should also continue. So I'm agreeing with your position. 572 MR. JANIGAN: Okay. But in this circumstance, the market premium belonged 100 percent to the ratepayers. 573 MR. PENNY: Times change, Mr. Janigan. Times change. 574 MR. JANIGAN: In that circumstance, is it not correct that this premium was going 100 percent to the ratepayers? 575 MR. POREDOS: In that particular circumstance, it went to the ratepayers, which was in the year 2000. 576 MR. JANIGAN: Okay. 577 MR. POREDOS: The year after, though -- 578 MR. JANIGAN: We'll get to that. 579 MR. POREDOS: We'll get to that? 580 MR. JANIGAN: We'll get to that. The reason it was appropriate at that time that Union submitted was that the allocation of rents was appropriate since infranchise customers had paid for the development of the storage. That hasn't change, has it? 581 MR. POREDOS: No. 582 MR. JANIGAN: Okay. And they pay for the ongoing cost of the storage pools as well as a rate of return on these storage assets to the shareholders? 583 MR. POREDOS: As those pools are put into rate base, yes, that's correct. 584 MR. JANIGAN: Okay. Now, in the RP-1999-0017 proceeding, and I've taken an excerpt from that and put it in these materials as well, on page 11 of these materials, page 134 of the decision, it's noted that Union -- if you found that reference. 585 MR. POREDOS: Page 11. 586 MR. JANIGAN: Yes. 587 MR. POREDOS: Which paragraph? 588 MR. JANIGAN: At the bottom of the page, 2.475. 589 MR. POREDOS: yes. 590 MR. JANIGAN: "Union also referred to the evidence of its witness, Ms. Elliott, who indicated that if transactional service revenues, storage and transportation, or long-term storage premiums were not available to Union, then it would have sought a premium or a growth factor under the cap." 591 Is it not apparent from that that the sharing of the long-term premiums in this case was important to offset the risk of PBR? 592 MR. POREDOS: In this particular -- 593 MR. PENNY: Sorry. This is -- you require some historic context for this statement, because my recollection was that the proposal was quite different. The 75/25 is the result of this case. 594 MR. JANIGAN: Yes. 595 MR. PENNY: Not Union's -- what Union's going-in position was, just to be clear. This is a recitation of argument on the positions of the parties. 596 MR. JANIGAN: Well, presumably -- 597 MR. PENNY: Not the Board decision. 598 MR. JANIGAN: -- presumably, though, Mr. Penny, the results of the case had something to do with what you presented, and in this case you presented this as the fact that you would need these -- some sharing of the storage premiums because of the risk of PBR. Is that not the case? 599 MR. PENNY: Well, I don't know whether there was any connection between the two or not. I do know that Union was asking for the closing of that deferral account as part of the total package that it was seeking in the introduction of performance-based rate-making. That's not what happened. 600 MR. JANIGAN: Well, I don't want to continue the colloquy. It's just the last part of that, and it can probably be addressed in argument, it seems to make clear my point. Let me push on. 601 MR. PENNY: It does seem like an argument point to me. 602 MR. JANIGAN: Now, further on in the decision, in paragraph 2.505, it's noted, and that's on the bottom of my page 18 of the decision -- of the decision page 141, 142, that: 603 "The balance of the long-term storage premium deferral account shall be allocated 100 percent to the ratepayers for 1999 and 2000 with incentive sharing for the long-term storage premium account to be effective January 1st, 2001." 604 Now, am I correct that 1999 and 2000 were cost-of-service years, and January 1st, 2001 is the start of the PBR? 605 MR. POREDOS: My understanding, and I can be corrected, was that the PBR, in fact, started at 2000. 606 MR. PENNY: 2001. 607 MR. POREDOS: 2001? Yes, sorry, then I would agree with your statement. 608 MR. JANIGAN: The original proposal was to start in 2000, Mr. Poredos, but was moved to 2001. 609 Thank you, Mr. Chair. Those are all my questions. I apologize for taking much longer than I had anticipated. 610 MR. SOMMERVILLE: Not at all, Mr. Janigan. 611 As I think I indicated yesterday that we'd break at 12:30, I think I misspoke. We will break now and resume at 1:00 -- actually, 1:10. That's to accommodate a conference call that Mr. Birchenough has waiting. 612 So we'll break now and come back at one -- 613 MR. PENNY: At 1:10. 614 MR. SOMMERVILLE: He's whispering in my ear that 1:00 will be fine. We'll come back at 1:00 and carry on for the rest of this panel's evidence. 615 MR. PENNY: Thank you, sir. 616 MR. SOMMERVILLE: Thank you very much. 617 --- Luncheon recess taken at 12:00 p.m. 618 --- On resuming at 1:05 p.m. 619 MR. SOMMERVILLE: Please be seated. Thank you. Are there any preliminary matters before we begin with Mr. Brett's cross-examination? 620 MR. BRETT: Mr. Chairman, I believe Mr. Aiken is next. 621 MR. SOMMERVILLE: I beg your pardon, Mr. Aiken. 622 Mr. Penny, I understand that Mr. Brown may be attending this afternoon to seek -- my understanding is, and I haven't heard from him directly, of course, but my understanding is to seek an extension of time with respect to the filing of evidence which was due today, and my understanding is that Mr. Brown is going to seek an extension of time to Tuesday morning. 623 MR. MORAN: That's right, Mr. Chair. Mr. Brown will be here about 2:30 so he can speak to it before you take the afternoon break. 624 MR. PENNY: Mr. Moran advised me of that, yes. 625 MR. MORAN: I left a message for Mr. Vegh in case he has an interest in the issue as well. So hopefully he'll get that and be here if he has an issue. 626 MR. SOMMERVILLE: Mr. Brett. I beg your pardon, Mr. Aiken. 627 I'm looking straight ahead at you, Mr. Brett, and I thought you were next. I beg your pardon, Mr. Aiken. 628 CROSS-EXAMINATION BY MR. AIKEN: 629 MR. AIKEN: Good afternoon, panel. My name is Randy Aiken. I'm here on behalf of the London Property Management Association and the Wholesale Gas Services Purchasers Group. You'll be happy to know, as I'm sure everybody in this room will be, that my estimate of an hour has been pared down substantially, and I hope to be substantially less than that. 630 My first question goes back to a response by Mr. Fogwill on the previous panel, and I'd asked him there to look at Exhibit C.1, tab 4. And at the bottom of the page -- and I don't think you need to look it up but it's on page 8 -- it says that a 1 percent variance in the heating degree day estimate represents a change in annual delivery revenues for Union of approximately 3.5 million, and he had updated that number to, I believe it was, 3 million, and then he referred me to this panel to confirm that. 631 So can you confirm that $3 million estimate? 632 MR. GARDINER: Mr. Aiken, the number is $2.7 million per 1 percent change in heating degree days. 633 MR. AIKEN: And what is the fixed charge that that estimate was based upon? 634 MR. GARDINER: There is no fixed charge in the calculation because we were varying the volume. So the unit rate for all service classes works out to about 7.4 cents per cubic meter. 635 MR. AIKEN: And is that using current rates in place for 2003? 636 MR. GARDINER: That is correct. 637 MR. AIKEN: So if the Board were to approve Union's proposal to increase the fixed charge from $10 a month to $14 per month, the corresponding variable charge would then decrease. Would it be possible to estimate what that $2.7 million would be under that scenario, as proposed by Union? 638 MR. GARDINER: I'm afraid I can't do that because that's outside of my area. 639 MR. AIKEN: Would that be something for the cost allocation/rate design panel? 640 MR. GARDINER: I believe so. 641 MR. AIKEN: Thank you. 642 I'd like to turn now to the first page of my book of materials 2, and I believe we'll need an exhibit number for that. 643 MR. SOMMERVILLE: I have the first volume. 644 MR. AIKEN: I have a very small one. 645 MR. SOMMERVILLE: Does Mr. Wightman have that? 646 MR. WIGHTMAN: Excuse me? 647 MR. SOMMERVILLE: Mr. Aiken's cross-examination materials, volume 2. 648 MR. WIGHTMAN: I apologize. 649 MR. SOMMERVILLE: Thank you, Mr. Wightman. 650 MR. MORAN: Mr. Chair, this will become Exhibit M.5.3, document entitled "LPMA/WGSPG Cross-Examination Materials 2." 651 EXHIBIT NO. M.5.3: DOCUMENT ENTITLED "LPMA/WGSPG CROSS-EXAMINATION MATERIALS 2" 652 MR. AIKEN: My first series of questions were going to be related to the other income category that's shown under operating revenue. Maybe I should ask, perhaps, Mr. Penny whether these the correct panel to deal with these other income items; and if not, what would it be? 653 MR. PENNY: The finance group are the people that deal with other revenue. 654 MR. AIKEN: So that would be Ms. Elliott? 655 MR. PENNY: That's right. 656 MR. AIKEN: Okay, thank you. 657 If you can go to page 2 of that material, and this ties in to the discussion earlier today between Mr. Janigan and Mr. Poredos, page 2 has a newspaper article from a month and a half ago, and I've highlighted a number of paragraphs there. My one question for Mr. Poredos on that is -- I guess maybe I should have a preamble to this. 658 My understanding of Union's position is that for 2003 and 2004, the S&T market has the potential to be weak, but in the longer term, there is potential for the S&T transactional services. 659 MR. POREDOS: Is that a question, sir? 660 MR. AIKEN: Yes. Do you agree with that? 661 MR. POREDOS: Yes. In terms of being weak compared to the recent past. 662 MR. AIKEN: Yes. 663 MR. POREDOS: Yes. 664 MR. AIKEN: And would you agree that the comments here by Mr. Tom O'Connor, the president of Duke Energy Gas Transmission, reflect that same sentiment? 665 MR. POREDOS: I haven't read this. I did read it a month ago or whatever, but I haven't read through the total. And your highlights don't show up on this photocopy so I'm not sure if you're asking me a specific question about one of his comments or... 666 MR. AIKEN: Okay, well, I'll read the specific area in, and this starts at the bottom of the second column, and it says: 667 "O'Connor said the acquisition of Union Gas Limited will allow it to make use of Duke Energy's extensive pipeline system which serves the east coast." 668 MR. SOMMERVILLE: Mr. Aiken, please be a little slower, thank you. 669 MR. AIKEN: "'It will allow us to have a more powerful product to offer customers on the east coast,' he said. 'It will make us much more competitive.' O'Connor said capturing a share of the east coast market will increase the value of the products the company offers - natural gas, storage and transmission. He said natural gas storage capacity in Lambton County is key to capturing new markets." 670 My question is: Does that basically reflect the same comments you provided earlier today? 671 MR. POREDOS: I don't believe that his comments are any different than our position. He was speaking broadly in terms of how the assets interconnect with the northeast markets, and I agree with that. 672 MR. AIKEN: Okay, thank you. 673 My next question is probably for you, Mr. Rogers. It has to do with the contract forecast. Have you had a chance to review the evidence filed in this proceeding of the Wholesale Gas Services Purchasers Group? It's Exhibit K.35. 674 MR. ROGERS: I have, but I'd like to look at it again. I have it now. 675 MR. AIKEN: Have you had a chance to review the rationale for the decrease in the M9 volume that is presented in that evidence? 676 MR. ROGERS: I haven't specifically looked at this in detail, so could you ask me a specific question on it, please. 677 MR. AIKEN: Do you accept the decrease in the M9 volumes, based on that evidence? 678 MR. ROGERS: Just let me refer back to something, please. Are these -- I'm going to need to -- my evidence aggregates the wholesale in my Exhibit C.1. I'm going to need to go to the disaggregated evidence behind that. 679 Mr. Reghelini, could you -- 680 MR. PENNY: Mr. Chairman, my recollection is, and we're just looking for it, that the question that Mr. Aiken has just posed was posed in an interrogatory and was answered, and that we'd indicated that we were going to update for that effect in the blue page of phase 2. But we just can't lay our fingers on it right now. 681 MR. SOMMERVILLE: Mr. Aiken, is that consistent with your recollection? 682 MR. AIKEN: Actually, sir, no, it's not. I don't remember that. 683 MR. PENNY: Well, I mean, Mr. Aiken should proceed, and we'll see if we can find it. 684 MR. AIKEN: The remainder of my questions have to do with the general service forecast and so I'll move on to those. 685 If you could turn up page 3 of Exhibit J.17.16. This is an IGUA interrogatory response, steps for weather normalization. Do you have that? 686 MR. GARDINER: J.17.16? 687 MR. AIKEN: J.17.16, specifically part G, the steps for weather normalization. Do you have that? 688 MR. GARDINER: Yes, I do. 689 MR. AIKEN: I basically understand steps 1 through 6, so I have two questions, and they're both based on step number 7. The first sentence says: 690 "Multiply the weather demand coefficients obtained from the forecast equations from the weather variances obtained in step 4 for each month." 691 Is step 4 really step 6? Is step 6 where the weather variance is calculated? 692 MR. GARDINER: Yes, it is, Randy. 693 MR. SOMMERVILLE: So should we correct our exhibit then? 694 MR. GARDINER: Yes. I apologize for the error. The reference should be to step 6. 695 MR. SOMMERVILLE: Mr. Gardiner. 696 MR. AIKEN: Then the second question I had on point number 7, again, the first sentence talks about the coefficients obtained from the forecast equations. Now, you have two equations; you have a volume equation and you have a use equation. Which of those two do you use, or do you use both? 697 MR. GARDINER: We use the use equation. 698 MR. AIKEN: The use equation, okay. This normalization methodology, is it a change from what the Board would have seen in EBRO 499, or is it unchanged? 699 MR. GARDINER: The mechanical pieces are the same. The weather coefficients have changed because in the EBRO 499, the use equation was based on double-log transformations, and in this multiple regression that you have in the blue page, there's no log transformation. 700 MR. AIKEN: Thank you. I want to turn to the price variable that's used in the various equations, and the forecast prices that you use are found in the response to Exhibit J.34.33 F, and I found them at tab 3.2 of that spreadsheet. I'm not sure if a hard copy was filed as part of that or whether it was just the information on the CD. 701 MR. GARDINER: The information was provided on the CD. 702 MR. AIKEN: Okay. Again, just a couple of questions here. My understanding is that these price variables, which are different for rate M2, rate 01 residential, as well as different numbers for the three classes of commercial and the one industrial class, are based on your March QRAM. 703 MR. GARDINER: That is correct, Mr. Aiken. The commodity component -- the commodity component of these burner-tip prices are what the customer pays at the retail home level includes the commodity portion, which is the QRAM, the current transportation charges, which are in our rate schedules; and the delivery portion of the rate in those prices are based on all the rate schedules going back in time. 704 So -- and let me correct myself a little bit, too. Also, when I was talking about the QRAM now, that was just for the forecast. For the history, we went back and picked up all the QRAMs from today back to 1990 and calculated the prices that a typical residential or commercial customer would pay, and we built it on the usage profile of the customer. So knowing how the customer consumed through the month, we basically put it through a billing calculation and came up with these prices. 705 MR. AIKEN: What would be the impact on these prices if they were updated to show the October 1st QRAM? Is it a small change, and if what direction? 706 MR. GARDINER: It's about three and a half cents. 707 MR. AIKEN: And which direction, up or down? 708 MR. GARDINER: Up. 709 MR. AIKEN: Up, okay. Roughly what percentage of M2 residential, commercial, and industrial customers are direct purchase? 710 MR. GARDINER: I don't have the exact percentage, but it's somewhere -- I'm going to say somewhere in the 35 to 40 percent range are direct purchase. 711 MR. AIKEN: How do you reflect the price that these customers are paying in this price variable for the gas commodity part? 712 MR. GARDINER: We premised on the system supply price. We did, in preparing the forecast, consult with energyshop.com to see what the prices would be, to see how our prices would compare. There are small differences, but since our time series was based on our own system supply price from the forecast, we continued to use our system supply. 713 MR. AIKEN: My next set of questions are going to be about the equations themselves, so the reference here would be Mr. Janigan's exhibit, Exhibit M.5.1 from this morning, pages 1 and 2. 714 Were these equations that were estimated tested for and, if necessary, corrected for autocorrelation? 715 MR. GARDINER: They were tested for autocorrelation. Where autocorrelation existed, there was no correction for it given that the capability of the equations to predict the 0.8 and the 1.4 that we were talking about this morning. Since I was quite close on the target, I didn't make the correction for autocorrelation in two -- in four of the 12 equations. All the other equations tested inconclusive. 716 From my perspective, this is not an issue because my forecasting tool, as we said this morning, the capability of the tool, based on the actual versus predicted as shown by the charts we were looking at, are performing very well for the residential and the commercial; and as I state in industrial, which is a challenging class, I'm still within 4 percent, or 4.4 percent. 717 MR. SOMMERVILLE: Mr. Aiken, both you and Mr. Gardiner seem to understand autocorrelation. I think I do, but I'm not sure that I do, and I wonder if you could give me a brief definition of what that is? 718 MR. PENNY: I think it would be appropriate maybe for Mr. Gardiner to do that, since he's the one who is testifying. 719 MR. SOMMERVILLE: I'd be happy to accept either illumination. 720 MR. GARDINER: Okay. Autocorrelation involves the residuals. When we set up an equation, we're trying to explain the use of the volumes, and we have variables, and there's a part of the volume that remains that's not explained, okay? That's the residual. You look at the residuals to see if there is a pattern. Ideally, they should be random. That means there's something that's not explained, it's random, and we can't associate anything with it. You can have, and I don't want to be too long on this, you can have positive and negative autocorrelation. And what it says is there is a pattern in the residual, something that's unexplained that has a pattern, and that would sort of suggest, well, can we find something to explain it. I didn't go that route. 721 As I mentioned, I thought that my forecasting tool was getting in within the 2 percent range, in some cases near 1, and I said to go that extra distance to autocorrect, which makes the equation much more complicated, much more difficult to implement internally, I said, I'll stay with my results instead. 722 MR. SOMMERVILLE: Thank you. That helps. 723 MR. AIKEN: Mr. Gardiner, did you test the equations and correct, if necessary, for heteroskedascity? 724 MR. GARDINER: No, I didn't. 725 MR. AIKEN: Did you do so for multicolinearity? 726 MR. GARDINER: No, I didn't. 727 MR. AIKEN: What is the impact on the equations of leaving in variables of questionable statistical significance on the other coefficients within that equation? 728 MR. PENNY: Well, is that a generic question at a level of theory? Because there is, insofar as I'm aware, no evidence that there are any values of a questionable nature in this evidence. 729 MR. AIKEN: That's a matter of argument, Mr. Penny. 730 MR. PENNY: No, it's a matter of evidence, Mr. Aiken. 731 MR. SOMMERVILLE: If I recall, I think, Mr. Aiken, you may be hearkening back to Mr. Janigan's line of questioning with respect to the reliability factor with respect to these equations, and there was a question that Mr. Gardiner answered where he indicated that, notwithstanding that they didn't meet that standard, he felt that the equations were still giving him coherent and credible results. Is that what you're looking for? 732 MR. AIKEN: That's correct. I'm looking at the individual T stats, not the overall equation. 733 MR. SOMMERVILLE: On that basis, Mr. Penny, I'm going to allow Mr. Aiken to continue. 734 MR. GARDINER: Can you repeat the question, Mr. Aiken? 735 MR. AIKEN: What is the impact on the equations you've estimated of keeping in the variables of questionable statistical significance on the estimate of the coefficients of the remaining variables in those equations? In other words, is there a bias imparted? Would the numbers change? 736 MR. GARDINER: I know that the weather variables, which are the dominant demand drivers in the equations, their coefficients did not change materially. I don't have the regressions. And I'm thinking back now as to whether -- I didn't do that type of exercise of, you know, running the regression with and without this variable. 737 I do know that when I was building these equations and testing different variables, and the price variable in particular, that my weather coefficients did not materially change, which gave me strength, gave me some confidence. 738 MR. AIKEN: Continuing on with the T stats for these coefficients that I think Mr. Janigan touched on. Can you provide the confidence level of each of the T stats on these two pages where the T stat is less than 2.0; and, failing that, could you provide the confidence level for the smallest T stat on these two pages, which I believe is a minus 0.11? 739 MR. GARDINER: Mr. Aiken, I'm not sure of the relevance of that. Obviously, both you and I, with a T statistic table, can obtain those confidence limits. As I mentioned earlier this morning, I fully recognize that a T statistic of .11 has a very low confidence level associated with it. The 1.35 -- and I may be speaking too fast and I should direct you to that table, but under residential rate 01 in the volume equation, there's a value of 1.35. I believe that would be at about 60 percent confidence. The .19 in the commercial, obviously it's going to be very low. But that's not the point. 740 I kept these price variables in because I wanted a forecasting tool that could provide management with an estimate, as prices change, especially two years ago, every week, as the prices changed in the winter, there's an e-mail, and what's the impact, what's the impact? 741 I wanted to have in our equations something that I could say, based on our consumption -- that could answer that question. So, you know, a table could be prepared, but I don't see the relevancy of it. 742 MR. AIKEN: Would it be fair to say, Mr. Gardiner, that the .19, .16, .11, those three in particular, have a level of confidence less than 50 percent? 743 MR. GARDINER: I agree. 744 MR. AIKEN: Okay, thank you. 745 MR. SOMMERVILLE: Just so I'm clear on this, too, do the fact that these numbers, the 1.35 that you referred to, Mr. Gardiner, as I understand it, that's a negative 1.35. 746 MR. GARDINER: Yes, and that's because in the calculation of the T statistic, it uses information from the gas price coefficient, which is also a negative, and you want -- well, the relationship between consumption and prices is inverse so it shows up as a negative. 747 MR. AIKEN: In the response, again, this is J.34.33 F in the spreadsheet from the CD, tab 5.2 is labeled "Discarded Equations, Commercial M2." Do you have that in front of you, Mr. Gardiner? 748 MR. GARDINER: One moment, please. 749 MR. SOMMERVILLE: This is J.34.35? 750 MR. AIKEN: J.34.33. 751 MR. SOMMERVILLE: I beg your pardon. Thank you. 752 MR. GARDINER: Mr. Chairman, Mr. Aiken is referring to information that was on that electronic CD. 753 And what was the reference, Mr. Aiken? 754 MR. AIKEN: Tab 5.2, "Discarded Equations, Commercial M2." 755 MR. GARDINER: Yes, I have it. 756 MR. AIKEN: The first equation I see there that was attempted, or was estimated but discarded, in addition, and I'm looking at the volume equations at the top, shows an additional variable of the mortgage rate with a positive coefficient. And I'm trying to understand why that was disregarded or that equation was not utilized. But my question centres around the next number of lines that says, "Goods producing EMP," I assume that's short for employment? 757 MR. GARDINER: That's correct. 758 MR. AIKEN: The service employment, total employment, and retail trade, each of those were tried in turn in the equation and they have positive coefficients, if I'm reading these tables correctly, which is what you would expect. An increase in employment would reflect economic conditions which would be reflected in higher commercial volumes. Do you agree with that, Mr. Gardiner? 759 MR. GARDINER: Yes, I agree. 760 MR. AIKEN: Now, the T statistics I see for those three or four estimates are, first of all, 0.96, 0.90, 0.99, and 0.86, and I also note that at the top of that page, the natural gas price is still with the proper sign, it's negative. My question is: If you're including variables with a T stat of .11, why did you disregard these equations when they have higher T stats? 761 MR. GARDINER: In some cases, if I take you to the bottom -- the use-per-customer equations, by adding some of those variables, my price coefficient became positive. And one thing that I didn't mention this morning is, if I had a relationship where I had a low T stat and the incorrect sign on the coefficient; in other words, higher price, higher consumption, which is not economically correct, I would not use it. So in some of the formulations you can see, especially with the use equations, I got the wrong sign. 762 MR. AIKEN: Yes, but I'm not asking you about the use equations. 763 MR. GARDINER: Okay. Rate M2, on the information that Mr. Janigan provided, shows a volume mean absolute percent error of 1.5, and the equation that you referred to that had the mortgage rate produced the 1.7; so with the equation that had service employment. There are others like the one that has total employment, it gave me the same. The other two, moving across to the right, gave me 1.6. I will concede that the goods employment one was 1.4, but I chose not to select it. 764 So when I looked through this, I was trying to keep a consistency in my structure between the use and volume in adding those variables; and on the use side I got improper relationships; on the volume side, I didn't get improvements. And that is the reason why I didn't -- I discarded those equations. 765 MR. AIKEN: If management came to you and asked you the impact on your average commercial customers of an economic slow-down, would you then be adding in something like a goods-producing employment number in the same manner that you added in the retail price impact when you were asked about the impact of price on use? 766 MR. GARDINER: The other point -- I'd like to, okay. The other point also in working with some of these equations is that these are provincial numbers heavily influenced by the city that we sit in right now, Toronto and the Greater Toronto Area. Most of our franchise territory is outside of this area. 767 So there was some concern about using -- there is concern I have, whenever I try and address that type of issue, because I've got this Toronto -- I could have this Toronto effect in that. In the other equations, I have weather and prices that are specific to Union Gas customers. 768 MR. AIKEN: Would this be a concern that you would have on the way that the customer attachment forecast is done, which I believe relies heavily on the Canada Mortgage and Housing Corporation forecast? 769 MR. GARDINER: We use the Canada Housing and Mortgage forecast, but we also, in preparing that, use the regional for Canada Housing and Mortgage which gives us the Toronto, Ottawa, and the Niagara region. So the remaining areas, as well as CMHC also shows urban areas in total so you get an idea of the rural. 770 So from that breakdown, we take that total provincial number and we boil it down to a number that is representative of our franchise area. So that's how we get rid of the Toronto effect in our customer-attachment forecast. As well, that's only an input to the channel management people. They take that, consider that. They also talk to builders and builders' associations in London and Windsor and the other major cities that we serve and, from that, build up the customer-attachment forecast. 771 So in that area I feel that we've addressed this Toronto bias, if you wish, in our customer forecast. As I was mentioning before, I'm concerned sometimes when I use economic data, you know, is it Ontario or is it the GTA that's affecting this? 772 MR. AIKEN: Those are my questions. Maybe I should go back to the contract question and have that as an undertaking, to see whether Union -- unless we've found an answer. 773 MR. PENNY: No, we didn't. I think the undertaking is the appropriate way to do it. 774 Can you just state clearly what it is that you wanted us to confirm? 775 MR. AIKEN: Yes. An undertaking to confirm that Union is accepting the decrease in the M9 volumes as reflected in Exhibit K.35 evidence of the Wholesale Gas Services Purchasers Group. 776 MR. MORAN: That would become Undertaking N.5.3, Mr. Chair. 777 MR. PENNY: Thank you. 778 UNDERTAKING NO. N.5.3: TO CONFIRM THAT UNION IS ACCEPTING THE DECREASE IN THE M9 VOLUMES AS REFLECTED IN EXHIBIT K.35 EVIDENCE OF THE WHOLESALE GAS SERVICES PURCHASERS GROUP 779 MR. AIKEN: Thank you, panel. Those are my questions. 780 Mr. Chairman, it's Friday afternoon before a long weekend so I'm going to ask you for permission to take my leave. 781 MR. PENNY: We should make him stay just out of spite, Mr. Chairman. 782 MR. SOMMERVILLE: You've heard Mr. Penny on the subject, Mr. Aiken. You're excused. Thank you very much. 783 Mr. Brett. 784 MR. PENNY: Can we just get that number. 785 MR. MORAN: Yes. It was N.5.3. 786 MR. SOMMERVILLE: Have a good, safe trip, Mr. Aiken. 787 MR. AIKEN: Thank you. 788 MR. BRETT: Thank you very much, Mr. Chairman and Mr. Birchenough. 789 CROSS-EXAMINATION BY MR. BRETT: 790 MR. BRETT: Good afternoon, panel. 791 I would like to start with you, Mr. Rogers. If you would turn up, please, Exhibit 34.38, or, rather -- yes, 34.38. That's a VECC interrogatory response, 34.38. 792 MR. ROGERS: Yes, I have it now. 793 MR. BRETT: That interrogatory response provides the level of gas consumption of the IPPs over the last four years. Now, Mr. Rogers, firstly, what you're referring to as the IPPs there, are they the 11 non-utility generators that have the long-term contracts with OEFC? 794 MR. ROGERS: Yes, that's correct. 795 MR. BRETT: All right. And that's all that that represents. 796 MR. ROGERS: That's right, yeah. 797 MR. BRETT: Okay. Now, could you tell me, and this may be in evidence and I apologize if it is, but what is the best estimate of the number of the bridge year for 2003 for the 11 non-utility generators? 798 MR. ROGERS: Just let me refer back to my blue-page evidence. I don't think -- I don't believe that number specifically has been broken out. It's incorporated in the explanations under the rate 100 and rate M7. 799 MR. BRETT: What reference are you making there? 800 MR. ROGERS: Sorry, I'm talking about Exhibit C.1, tab 2, page 12 of 12. 801 MR. BRETT: Right. That's the white pages? 802 MR. ROGERS: Yes. 803 MR. BRETT: Could you tell me, do you have an approximate number for that 2003 consumption; or failing which, perhaps I could get an undertaking to provide that, if you'd prefer that. 804 MR. ROGERS: Yes. I think that's the preference. 805 MR. MORAN: Mr. Chair, that's Undertaking N.5.4, an undertaking to provide the 2003 consumption by the 11 non-utility generators? 806 MR. BRETT: Yes, that's right. 807 UNDERTAKING NO. N.5.4: TO PROVIDE THE 2003 CONSUMPTION BY THE 11 NON-UTILITY GENERATORS 808 MR. ROGERS: What I can do, or would like to reference is the -- and we will certainly do that. But if I could just, in the way of -- it's probably better to do it in the undertaking. 809 MR. BRETT: You may be anticipating, in part, my next question, Mr. Rogers, because my next question is: We don't know what that number is, but whatever it happens to be, can you tell me what decline from the 2003 number for the 11 NUGs gas consumption you are forecasting for 2004? And just by way of elaboration, I suspect that is where your reference to rate 100 in Exhibit C.1, tab 2, page 12 comes in. 810 MR. ROGERS: That's correct. 811 MR. BRETT: That number there in rate 100, gentlemen, if I can just state this to hasten things along, but opposite Ontario power markets, it shows a decline of 357.4, 10(6)m3. Mr. Rogers, that 357.4, that's the estimated reduction -- and I must correct myself, that's the estimated reduction between 2004's forecast for the 11 NUGs consumption and the 2002 statement, is that the 2002 amount? 812 MR. ROGERS: That's correct. As I've described before, Mr. Brett, that forecast was done in the summer of 2002 and, in fact, was reflective of the probability that these contracts would be purchased down by the financial arm of the newly formed open market. And so what the reflection was here was that that action would be taken. In the blue-page update, which I would reference, page 5 of 5, you will see -- sorry, Exhibit C.1, tab 2, page 5 of 5 updated, or blue page, you will see under the rate 100 that, in fact, we have, for 2-04 blue versus 2-04 white, added back in 129. 813 MR. BRETT: Okay. 814 MR. ROGERS: So what the reflection is there is that we now know we have protection through to the end of November of this year for that level of contracted demands. 815 Now, we did not return the entire 359 that we moved on the white page, and the reason for that was that we still believe that they will burn down to their contractual minimums and I've already put that on the record in terms of the gas-pricing situation, that the 2-04 gas price is expected to stay up at least the 2-01 differential in pricing for all fuels and it will, in all likelihood, be sold off to the market, so that is why the number is not returned in its entirety. 816 MR. BRETT: Thank you. I recall that from yesterday's transcript. And so the 129 is added back in in the blue-page update so that the resulting reduction, estimated reduction, if you like, from 2002 is the 357.4, or I guess you're calling that 359, less the 129; right? 817 MR. ROGERS: That's correct. 818 MR. BRETT: And then we won't know what the reduction is from 2003 until we actually see the 2003 number, but the same exercise can be done. 819 MR. ROGERS: Yes, it can. 820 MR. BRETT: Okay. Now, your comment there about your contracted capacity leads me to my next question, which is that -- I'd like you to turn up, if you would, just so we have a transcript reference, paragraph 1311 of yesterday's transcript. That's Volume 4. 821 MR. ROGERS: Okay. 822 MR. BRETT: Yesterday's transcript. And it's paragraph 1311. 823 MR. ROGERS: Just give me a moment and I'll do that. I'm there. 824 MR. BRETT: The reason I do this is I just -- you will see there, you said yesterday, and this is, I think, referring to what you just said a moment ago: "You will find that we have adjusted them back up because, in fact, we know now that we have contractual protection through 10 months of 2004, and we only are posed for the last two months of 2004." 825 So what I take you to be saying there, Mr. Rogers, is that you have contracts with these NUGs to provide them with distribution service up until November 1 of 2004 -- 826 MR. ROGERS: Right, this is to the minimum contractual commitment to burn to the end of 2004, that is where there are only two months left. 827 MR. BRETT: Okay. 828 MR. ROGERS: Now, the differential we've just talked about between 359 and 129 is made up of two component parts, the 129 which we have returned, and the remainder, the balance of the difference between those two is, in fact, the increase above the minimum contractual burn that was experienced in 2002. We have removed that because we do not believe that same opportunity that arose because gas prices were lower in 2002 will repeat itself for 2004. 829 MR. BRETT: All right. 830 MR. ROGERS: So there are two distinct parts. 831 MR. BRETT: And the 129 represents the two months' worth of -- 832 MR. ROGERS: It actually represents the 10 months that -- it represents -- 833 MR. BRETT: The months during which you're protected? 834 MR. ROGERS: Right. The last two months, that's correct, November, December. 835 MR. BRETT: And just to clarify, what you are saying is your firm contractual obligations that you have from the NUGs just simply extends to their minimum contract burn requirement under their power contracts, and then anything above that, you don't have a firm contract with them in any event. 836 MR. ROGERS: That's correct. This is a discretionary burn on their part. 837 MR. BRETT: Okay. Now, with that by way of background, I have just a few questions to put to you. The first is: Are you aware, Mr. Rogers, that a number of the NUG contracts, now, these are the long-term contracts between the 11 non-utility generators and the Government of Ontario in the person of the Ontario Energy Finance Corporation, are you aware that a number of those contracts -- under a number of those contracts, the non-utility generator does not have the right to sell its gas to a third party; in other words, it must produce or take the gas in any event? It doesn't have the right to sell to third parties. Are you aware of that? 838 MR. ROGERS: I don't know that. I'm not privy to the contracts. The only issue I would ask for a clarification, because I have not read it, is it in the event that now IMO versus Ontario Hydro calls on that plant to burn, I could see that to be true. I would question if, in fact, they do not need that plant to run because they have other cheaper disbatchable options that they would force that contract to run, hence the opportunity may be available to them if IMO does not call on them to be run to be free of that commitment and sell the gas. I am not privy to it, but that would be my opinion. 839 MR. BRETT: All right. Well, I'll come back to that question of dispatchability in a moment. Just to focus on this level of -- operating level that the NUGs are likely to operate at in 2004, because I think you'd agree with me this is the nub of the matter. 840 First of all, just as a clarification, you're not suggesting that any of the NUGs won't operate in 2004, that they'll disappear or pass out of existence or close down? 841 MR. ROGERS: What I'm suggesting in my blue-page update is that we have contractual protection for 10 months of 2004. We do not for the last two. I have no way of knowing whether OEFC will enter the equation and buy them down. 842 MR. BRETT: All right. And you're aware, are you, that many of the NUGs are cogenerators and therefore they produce steam as well as power? 843 MR. ROGERS: That's correct, a small amount of the steam given to the host. 844 MR. BRETT: And that they must, under their steam contracts, in order to honour their steam contracts, which are typically 20-year contracts, long-term contracts, with an industrial host, they must continue to operate at a certain a level simply to ensure -- they must operate at such a level that one and the same time allows them to honour that steam contract and also not destabilize their equipment. 845 MR. ROGERS: Right. My understanding in the vast majority of these, I'm not privy to the exact number, but the number they would have to meet the contractual commitment they have with us far exceeds any of the commitments they would have with to steam host to deliver the steam, which typically is the lower pressure because it's the second-stage reduction left over after the production of electricity. 846 MR. BRETT: You're aware also, are you, that to look at their contracts with -- the NUGs all have contracts with OEFC, and under those contracts, I think you'd agree with me, they have minimum requirements to operate. If they don't meet the minimum requirements, they will be facing penalties. 847 MR. ROGERS: I don't have -- 848 MR. BRETT: Let me pose my question, if I could. Are you aware that those contractual obligations to operate are typically in the range of 80 percent; in other words, they are high obligations, before they meet a capacity. And if they fail to operate at that level, they get penalized. They are typically in the 80 percent range. 849 MR. ROGERS: I'm not privy to the contracts so I can't tell you that. What I do know is they can also run the supplementary boiler and not the turbine, which uses most of the gas, to fulfil most of the lower pressure requirements at that facility. So I do know that much. What I don't know, Mr. Brett, are the details of the contract. We don't own any IPPs and hence aren't privy to a contractual, inside understanding. 850 MR. BRETT: Have you considered the fact in your estimates that some of the NUGs, not all of them, but some of them are negotiating with the OEFC to actually produce somewhat more power next year because they have certain incremental capabilities that they have not been able to -- under the existing contracts, not been able to produce because they have a ceiling that they have to pay for? 851 MR. PENNY: Can I have clarification on that, Mr. Chairman, whether we're talking about non-utility generators in the province of Ontario generally, or some or all of the 11 that are in Union's franchise area? 852 MR. BRETT: I think we're talking about -- well, I know we're talking about some of the 11 that are in the Union franchise area. 853 MR. ROGERS: I think the way I would answer that is I'm not privy to that, but I would return to my point that we have added in 1,220 megawatts of brand new, large-scale, high-efficiency cogeneration, and added 1,068, 10(6)m3 to our forecast. That would compete to any option that would be offered up by these particular individual generators. And so if those generators were successful by discounting because they have some cost or some way to do this, then the number that I put if nor the generators that I have would possibly be displaced. 854 MR. BRETT: That brings me to this question of the dispatchability. You have stated in your -- you mentioned this question of them being displaced. 855 MR. ROGERS: Can you give me a reference? 856 MR. BRETT: Yes. It's C.1, tab 2, page 6 of 12. That's the white-page evidence. 857 MR. ROGERS: Thank you. 858 MR. BRETT: I'll just read the sentence out. 859 "There continues to be considerable uncertainty about the operations of the 11 IPPs located in Union's franchise area, particularly with respect to increased dispatching by the IMO in favour of less costly electricity supply alternatives." 860 Now, my understanding, Mr. Rogers, that I put to you is that these 11 NUGs are -- they have established protocols with the IMO which effectively are self-scheduling, gives them a self-scheduling status, which means, in practical terms, that they are always dispatched. They don't bid; they don't have to bid into the dispatching mechanism. They're always running. The only caveat to that is that they're producing the power and they're producing the power within the -- within the ambit of the contract they have with the OFC. But so long as they're doing that, so long as they're not trying to produce more than the contractual capacity ceiling they're entitled, there never is a question of them being dispatched. They have self-scheduling status which was negotiated with the IMO and they always -- their power is always fed into the market. They're price-takers, regardless of what the market price is. 861 Now, if you differ from that, I'd like you to tell me your reason for that. 862 MR. PENNY: Well, Mr. Brett, you're not here to give evidence. I think the proper question is to determine whether what all -- the facts that you were just trying to put on the record are, in fact, known to Mr. Rogers at all. 863 MR. BRETT: Let's start with -- let's start by asking it that way, Mr. Penny. You have a point. 864 Is that your understanding of how it works with the 11 NUGs? I'm not talking now of the TransAltas of this world or the new merchant plants. 865 MR. ROGERS: I heard a statement, not a question, Mr. Brett. Could you pose it as a question? 866 MR. BRETT: I just made it a question. I said do you agree with the description that I gave of you how this works? 867 MR. ROGERS: I'm trying to remember. It was quite long. 868 MR. SOMMERVILLE: I think what Mr. Brett was suggesting was that the cogeneration plants that he's been referring to are not subject to a dispatch queue, if you like, from IMO, provided they are operating within the contractual context, the amounts that they're contracted to produce. That in that case, they are not dispatchable, per se, they simply operate. 869 MR. ROGERS: Right. Thank you, Mr. Chairman. What I would answer to that is that that is the way that we would set the contract demands with these customers, so that the commitments we have from them are then covered except for any incremental business that would happen from increasing the load factor that is originally predicted. So at the amount that they are contractually committed to burn, it's my understanding that lines up with our minimum contractual commitments. If, in fact, they are successful, for instance, in getting on to -- because these are not 24-hour operations, at least that's what they tell us when we design the rates. And if they go outside what I believe is the normal 16-hour operation and, in fact, are able to or are to be called on during certain times, either at peak electricity needs or if they have cheap electricity to offer, that they would operate more and the load factors would, in fact, increase. And that's the way I understand that it works. So hopefully that helps. 870 MR. BRETT: It does. I think we're roughly on the same track, Mr. Rogers. I think you'd agree with me that there is a distinction between their contracted requirement to burn and the total that they might burn in certain circumstances. They might very well burn more than that, and that would probably depend on some pricing relationships between the price of electricity under their contract at off-peak hours and the price of gas; correct? 871 MR. ROGERS: Yes. And remember these are demand commodity, so the demand portion would cover a large part. Any of the commodity volumes that would be added would be at a lower price and the load factor would enter in, so I agree. 872 MR. BRETT: So you agree with me that they're not subject to dispatch -- I think you've agreed that far -- but you've said essentially that, yes, that's right, they're not dispatchable; but on the other hand, we have set our goals with their minimum contract burn obligations in mind. 873 I guess the only remaining question is: You'd also agree with me that they are not subject to dispatch in the event that they produce more than their minimum contractual burn. As long as they produce within their overall contract with the OEFC. 874 MR. ROGERS: I would agree with the caveat that the experience we've had during 2001, which demonstrates this, and it's shown in the interrogatory, is that even if they have an opportunity to do that, there tends to be a higher opportunity when gas prices run up the way they have now, to sell the gas and make more money. 875 Now, I'm not privy to a restriction and whether they can or can't. Again, I would go back to, if it's an opportunity burn, that they're discounting to get out on the IMO dispatch above their contractual minimums, then they may burn. I would offer that my educated guess is that they would take one of two doors; either to discount and dispatch and make electricity above and beyond their minimum amounts, or take the gas that they get 365 days a year, which is a good price because they locked it in at the beginning of the 20-year period, take that out into high gas-price markets and sell it off if that price exceeds what they could get by dispatching incremental discretionary electricity. That's the basis for our point. 876 MR. BRETT: I think, having said that, you'd agree with me that if they weren't able to sell the gas under their contract to a third party and they had some fixed obligations, their marginal cost of producing additional power might be quite low, so they would be likely, in that instance, to produce more power. 877 MR. ROGERS: And our experience was in 2001, which is the similar situation, they did not, they sold the gas. 878 MR. BRETT: Some sold the gas, sir. 879 MR. ROGERS: Okay. Some sold the gas. 880 MR. BRETT: All right. Just to make sure that I'm sensitive to the timing here, I'm going to move on to a few questions on the demand on the non-contract section of this. Just let me check my notes for a moment. 881 MR. SOMMERVILLE: We're not running a race, Mr. Brett. 882 MR. BRETT: All right. Thank you, sir. I do want to honour my obligation, or come very close to it. 883 MR. SOMMERVILLE: Thank you. 884 MR. BRETT: Just a couple more questions for you, Mr. Rogers, of a general nature. 885 The 76 largest customers that you speak of in your evidence -- and, sir, I honestly don't know that you need to turn this up. I think you, at some point in your evidence, made the observation that for the very largest customers, you have a sort of personalized way of -- you go out and discuss with each of those large customers what their likely gas consumption is going to be. 886 MR. ROGERS: Exhibit C.1, tab 2, 3 of 12. 887 MR. BRETT: Right. And now my question is, sir, are these various power plants, the 11 NUGs that we've spoken of, the three new merchant plants, the TransAlta, Atco, OPG, and Imperial Oil, the Lennox plant, are those 15 entities included in the 76? Do you give them the same individual treatment that you give the 76 large industrials? 888 MR. ROGERS: The reason I hesitate, I'm not sure whether we got the Imperial contract in to the 76 number, the white-page number. But subject to being close, the answer is yes, we do go and see those people because they are in that category of what we call our largest industrial and power customers. 889 MR. BRETT: Can I ask you, you obviously go back -- you say you go and see them and discuss the requirement. You then come back and form a view as to what your forecast should be. Do you then tell them what your forecast is going to be for them? And I guess coupled with that, perhaps while you're -- in the event that they -- if you do tell them what the forecast is going to be for their individual facility for the test year and they disagree with you and say, Mr. Rogers, we think that you're 10,000 units too low for us or 10,000 units too high. If you have a dispute with them, what happens, really? 890 MR. ROGERS: Well -- 891 MR. BRETT: First of all, sir -- I'm sorry, I didn't mean to pile questions on you. 892 Do you go back and tell them, first of all, what their forecast is going to be for their facility? 893 MR. ROGERS: We don't go back and give the specific forecast out to each individual customer. And the reason we do that, we take their input, we look at the historical patterns, depending on how many years they've been in play, and look at the number historically, talk to them specifically about any changed circumstances that they can nail down in terms of whether they are, for instance, going to have a permanent plant closure, are they going to have a permanent addition, is there a significant shift in what the operation will look like, and we will adjust that accordingly. And we are very close to these customers so that it's an aggregate of that judgment of the manager who manages each of these individual sectors. 894 If we go to the remainder of my market, which is roughly 600 customers, we don't do that specifically. We do it more on a historical basis, and we do incorporate any specific growth or closures that we know about. But with the large customers, we work together to generate a forecast. 895 MR. BRETT: And I'll go back to those 600 just briefly in a moment. But for the large ones, the 76 plus the power producers, have you had, on occasion, disputes with them over whether or not -- well, I'm sorry, you're saying you don't tell them so they don't really know. Once they've had their conversation with you and you've had the back and forthing, they may not really know what your number is for them. I think that's what you're saying to me. They don't know the individual number. 896 MR. ROGERS: I don't administer these contracts personally with each and every one of the accounts. If the account wants to know what the number is, we don't have a policy against giving the person the number. What we don't do is put that out and publish it in the public domain. 897 MR. BRETT: Right. 898 MR. ROGERS: Specifically, we have never reached an impasse with a customer where it is elevated beyond an ability to agree. And if we take out the factors that I've already put on the record, we come in very tight to the numbers in most of my contract areas. It is the power area that is the most difficult to forecast for the reasons I've already put on the record. 899 MR. BRETT: And for someone in the power area, if they had a concern and they had a -- I guess the way the system works now, if they weren't happy with what they thought their individual; forecast was, if they had some way of finding out, their remedy would be to come here, I suppose, during the rates case. That's their remedy at the moment. I mean there's no private dispute resolution process or anything of that sort to sort out that kind of a thing, that sort of a difference offline, if you will. There's the rates case. 900 MR. ROGERS: In the case of the IPPs, for instance, with most of that being firm, it's a rate-setting issue as opposed to anything else, because the CD, because of demand components -- level. And if we're debating over what level of load factor you're going to have in between, then it's just a matter of the commodity component. So the rate's in place and the... 901 MR. BRETT: And what do you do -- for the other 600 or so, you don't have the equation approach that Mr. Gardiner is talking about. You have a general historical comparison. You look at what they've used and you make a judgment based on economic -- 902 MR. ROGERS: We do have a group that interacts with the top roughly 125 to 150 of those customers because they're larger than the remaining 450, and anywhere where we specifically know there are closures or additions or changes, we would add those in. The other ones, we are fairly homogenous and fairly easy to predict and repeat themselves. So we don't spend the time or the money or have the manpower to go down into that level. 903 MR. BRETT: Thank you, Mr. Rogers. 904 Mr. Gardiner, on the question of reasonableness, this is a test that you perform on each of the seven NAC forecasts; is that correct? 905 MR. GARDINER: That is correct. 906 MR. BRETT: And you've gone through, in some detail with Mr. Thompson, how you do that, the mechanics of it and I won't get into that. But in the white -- you start off, I gather, with -- and I'm looking at your white page here, C.1, tab 1, page 7 of your evidence. When you first make your forecast, you're making the forecast for 2004, you're making for -- you're making in 2002, as you've said. And so this reasonableness test that you have is that you look at -- well, you explained it yesterday, you look at what's happened in 2002 so far and then you make your adjustment, if that is necessary. Is that roughly right? I mean I don't need you to describe in detail again what you do -- what you did. But if I'm -- 907 MR. GARDINER: I believe we have it. I think we're right. With the blue-page evidence, okay, which was completed in May/June of this year, I had the first three months of this year. 908 MR. BRETT: And did you look at those first three years to form your outlook? 909 MR. GARDINER: That's what I did. 910 MR. BRETT: Okay. That's one of my questions. And then I guess the first question really is -- well, let me reverse the order of these. If you do that in the way you've described, those two different time periods, and then you say you move that forward, you effectively take that result for 2003 and you apply it to 2004, and my question is, why would you do that? Why would you assume that what would be appropriate in 2003 would necessarily be appropriate in 2004? 911 MR. GARDINER: There's an assumed -- it's assumed that the error or the -- being off the mark, if you wish, missing the net, in 2003 would continue in 2004. 912 MR. BRETT: All right. And my second question is: Why would you do the reasonable test at all? And the context for my question is you've got these two econometric equations that you use. You put a lot of effort in trying to hone -- hone down the variables, the weather, the efficiency, the price. You've discussed those things with others. Why, then, do you have this reasonableness test? Why is it necessary or desirable? 913 MR. GARDINER: It's desirable because I've found that, by doing it, I'm able to assure myself, first of all, that I'm on -- that I have a forecast that's on track. I've found that it's helped make the forecast more accurate, because I'm always concerned that, for some reason, I go through this whole process, I have an estimate, and this is my last piece of information to tell me is the estimate going off track. I have no other -- because I have no other information. It's a litmus test. 914 MR. BRETT: But, sir, with respect to the -- I can see that with respect to your best estimate of the -- I think I can see it with respect to your best estimate of the bridge year, 2003, but it only performs that function -- would you agree with me that it only performs that function with respect to the forecast if you assume -- if your point about carrying forward the reasonableness adjustment is valid? 915 MR. GARDINER: Essentially the adjustment is -- I don't know what's causing -- if we have a certain rate class that's outside of the range, I don't know the reason for it, so I make the assumption that the unknown takes place one more year. As I mentioned earlier this morning, the adjustment is very small. I make the adjustment on either side; it can be a positive adjustment or a negative adjustment. So I'm not biasing anything. It's because there's something, and to me it appears I'm off track. If I looked at two standard deviations around the equations, I feel I would -- I would not be as accurate. I've imposed a fairly tight sort of test. 916 MR. BRETT: Can I ask you just with respect to the different treatment of -- the treatment that you gave -- the adjustments that you made under the white-pages evidence and then under the blue pages. On the white pages, this is at your white-page evidence, C.1, tab 1, page 8, you say, line 12: 917 "For the residential and commercial groups," this is in the white pages, "no adjustment was needed as the year 2002 NAC forecast was within one standard deviation of the outlook," sorry "than the 2002 outlook." 918 And then at the blue pages, C.1, tab 1, page 2, you say: 919 "Five of the seven --" I'm reading from line 16: 920 "Five of the seven NAC values have increased in this update. In this forecast, two of the seven NAC estimates were adjusted for the reasonableness test. The adjusted NACs are residential M2 and commercial R10. Both NACs were adjusted downward." 921 So can you explain to me -- I guess two questions: 922 One, what were the -- this, you may have answered yesterday, and if you have, I apologize. What was the M2 NAC before you adjusted it downwards, before you employed the reasonableness adjustment? And ... 923 MR. GARDINER: In the blue pages, the adjustment for the residential M2 group was 14 cubic meters. 924 MR. BRETT: And where is that -- where are you taking that from, sir? Is that a -- you're quoting me a piece of the record there? 925 MR. GARDINER: That's from my notes, but we also have it in the interrogatory. 926 MR. BRETT: If it's in your notes, that's fine. 927 MR. GARDINER: There was an interrogatory that Mr. Aiken helped me with yesterday. 928 MR. BRETT: Yes. I think I recall that exchange. 929 MR. GARDINER: I just found it right now, Mr. Brett. 930 MR. BRETT: All right. 931 MR. GARDINER: Go to J.34.34. 932 MR. BRETT: Right. 933 MR. GARDINER: Page 2 of 2. 934 MR. BRETT: Right. 935 MR. GARDINER: And you will see in the right-most column under "Base case NAC" -- 936 MR. BRETT: Page? 937 MR. GARDINER: Page 2 of 2 of J.34.34. 938 MR. BRETT: I'm sorry, I have it now. 939 MR. GARDINER: Okay. If you look underneath the heavy bold total "residential rate M2" on the right hand. If you look at the -- 940 MR. BRETT: Of 2578? 941 MR. GARDINER: Right. You see the small minus 14s. 942 MR. BRETT: I see it now. 943 MR. GARDINER: Those are the adjustments. As well, on the commercial rate 10, for some reason that title went to the left -- 944 MR. BRETT: That should be in the blank there. 945 MR. GARDINER: Yeah. 946 MR. BRETT: Okay. Fine, thank you. 947 Now, could you just -- is there anything -- and I guess you would say that the reason that you made the adjustment the way you did here, but not in the white estimate, is that your basic blue-page estimate was different than your basic white-page estimate. 948 MR. GARDINER: That's correct, Mr. Brett. When I did the exercise the previous year, the same methodology, the same litmus test, those service classes passed the test. 949 MR. BRETT: So you're saying they passed it and they passed it in the white page -- 950 MR. GARDINER: But not in the blue page. 951 MR. BRETT: -- but not in the blue page. And the reason they didn't in the blue page is, again? 952 MR. GARDINER: For that small adjustment, I do not know. 953 MR. BRETT: Well, that, I guess, was my question. I wasn't sure. It was a little bit convoluted, but I take the point. 954 All right. Just very quickly going back to 34.34, J.34.34, I just want to make sure that I understand how the -- and I'll stick with the M2 for consistency, but how these numbers fall out. Well, I guess we'd better do the residential rate 01, the one above it, because that's what my case study here shows. 955 MR. GARDINER: Okay. 956 MR. BRETT: On the right-hand column, "base-case NAC," you have a number N/A 2,637. 957 MR. GARDINER: That's correct. 958 MR. BRETT: And then under equation 1 and 2, you have respectively for 2004, 2,673 and 2,600. Now, first of all, do you add those two and divide by two to get the 2,637. 959 MR. GARDINER: That is correct. 960 MR. BRETT: To get from the 2,367 to the 2,642, which is the final number, I gather, how does that work? Well, two questions, Mr. Gardiner: First of all, what happens to the 48 under reasonability under equation 2? 961 MR. GARDINER: As I mentioned yesterday, the format of this table was to respond to the IR of what would it look like if you had gone through and done the forecast on the basis of each equation. 962 The reasonability test in the blue-page evidence is performed on the base-case NAC. I just may want to clarify something here. As I mentioned yesterday, that 2,637 is the NAC that reflects the fact that we added the minus 32 for DSM, added the 40 for the marketing plan, and subtracted that small adjustment for the technical standards change on hot water tanks -- 963 MR. BRETT: If I could just stop you there. Those computations are in the 2,637, or do they explain -- 964 MR. GARDINER: Yes, they are. 965 MR. BRETT: They're in the 2,637. 966 MR. GARDINER: Yes. 967 MR. BRETT: What about the 48? Is that something different altogether? 968 MR. GARDINER: That is something different altogether. 969 MR. BRETT: That's the reasonability -- 970 MR. GARDINER: If we had only used equation 2. 971 MR. BRETT: Right. Oh, I see. If you'd only used equation 2 and hadn't used equation 1, you would have made a reasonability adjustment of 48. 972 MR. GARDINER: Correct. 973 MR. BRETT: But you always use the two equations. 974 MR. GARDINER: Correct. 975 MR. BRETT: So why do you still have that there? 976 MR. PENNY: That's what the question was. 977 MR. BRETT: I'm sorry. And how do you then get from 2,637 to 2,642? It's a small adjustment, but just in terms of understanding the methodology. If everything -- if things along the -- I had originally thought the three -- 978 MR. GARDINER: Maybe I've got -- 979 MR. BRETT: Is it the case, perhaps, that the three computations on the left-hand column are actually netted out and you get something very close -- 980 MR. GARDINER: Yes, that's correct, and I apologize. 981 MR. BRETT: All right. That's fine. 982 Now, the only other subject I wish to ask you about briefly, passing from reasonableness to the way that you handled energy efficiency, and there has been some discussion of this. I think I have just very few questions on this. The first is a stand-alone question. 983 In your -- I'm sorry, I guess this would be Mr. Rogers' question. Mr. Rogers, if I can flip back to you for a moment. In your white-page evidence, C.1, tab 2, page 12, in the table, you have put an energy efficiency per year factor in there of 2 percent under each of the rate schedules there, each of the seven schedules, so you've got it across the board, 2 percent industrial efficiency per year, and that will include the power plants, it includes everything -- I guess it includes everything, pretty much. 984 First of all, am I right that that's across the board? 985 MR. ROGERS: No. It's on the industrial accounts. We made the judgment call that the Independent Power Producers were new, that they were not going to be subject to efficiency improvements because, in fact, the 11 IPPs and the new cogens that are coming on now are, in fact, highly efficient. So we adjusted the number and only took into account the industrial portion. 986 MR. BRETT: Now, you'd agree with me -- I'm not challenging what you're telling me you've done -- you'd agree with me though that the way these numbers are displayed here though, you haven't made that distinction? 987 MR. ROGERS: Yes. Just as soon as I looked at it, I understood that. 988 MR. BRETT: Can I get an undertaking for this to be revised -- 989 MR. ROGERS: I think if we can clarify that this is, in fact, correct, that this is only attributable to the industrial loads under that category, that will be the clarification. 990 MR. BRETT: All right. Well, if you're clarifying it now, I guess that's all we need, you're saying? 991 MR. ROGERS: Yes, that's my statement that that's how it's calculated. 992 MR. BRETT: Fine. I'm satisfied with that. 993 MR. SOMMERVILLE: Thank you, Mr. Rogers. 994 MR. BRETT: Now back to you, Mr. Gardiner. At 17.16G, I think is the reference, you've been asked about -- this paragraph, for some reason, has gotten a lot of attention, apparently. You may have been asked about this. 995 I wanted to ask you about the 17.16, paragraph G9, and I'm just going to read this to give everybody a bit of context. You're saying -- 996 MR. SOMMERVILLE: Slowly, Mr. Brett. 997 MR. BRETT: Sorry. 998 "The contention that NAC is likely to decline in 2004 arises primarily from the energy efficiency assumption. The energy efficiency index is a significant demand driver in the econometric demand forecast equations. Energy efficiency has increased over the last 12 years and is forecast to continue to increase. On average, this decline due to energy efficiency is about 24 cubic meters per year." 999 What is the basis for that, Mr. Gardiner, that 24? And coupled with that, is this the case for just M2? Is this addressing just M2, or is that the case for each and every one of the seven rate classes or rate service groups that you make the NACs for? 1000 MR. GARDINER: If my memory serves me right, it's a residential rate class. 1001 MR. BRETT: Okay. And back to you, Mr. Rogers. On the efficiency rider of 2 percent that you've placed on all of the industrial sector, and I'm right in saying that you have not differentiated -- I mean, that 2 percent is applied to every single contracted industrial customers, all 600 plus 76 of them. 1002 MR. ROGERS: That's right. 1003 MR. BRETT: You don't distinguish between pulp and paper, petrochemicals, auto manufacturing, warehousing, and so on? 1004 MR. ROGERS: To be clear, this explains the efficiency differential for each of them. Now, in the large accounts, there are specific project adjustments, but we didn't try to do that across the entire population or we would have driven ourselves round the bend. So this is -- 1005 MR. BRETT: But you're saying that, for example, if you had one of your major clients, a paper mill, and did a large heat recovery project, you would factor that in to your forecast for the relevant year? 1006 MR. ROGERS: Yes, we do. That would constitute a reduction in their use and emergency usage. And I think, there's an undertaking that we described how -- what we have in terms of backup for that and how we use federally, cross-Canada, the developed data to substantiate that. In fact, the 2 percent that we use, if you look at that table and bring it up in the recent years, the numbers are higher, on average, they're more like 4 to 5, driven by higher gas prices. So we believe the 2 percent is probably conservative in terms of its level. 1007 MR. BRETT: Sorry, Mr. Rogers, you're going to provide an undertaking response along those lines? I'm familiar with that table that -- I'm familiar with that information from the federal agencies. It was the -- what caught my attention was your -- well, let's leave it at that. I think that you've answered my question. 1008 MR. ROGERS: The table speaks for itself. 1009 MR. BRETT: All right. Just to be doubly clear on that. This 2 percent factor is not based on any measured energy efficiency results that you've done on -- there's no -- 1010 MR. ROGERS: We take into account anywhere that we know there are energy efficiency projects that are in place. But what we have done here is reflected, as an explanation for our variances, 2 percent across to each of the rate classes, as that is what an average efficiency improvement is, according to the data and the table and the backup we have. 1011 MR. BRETT: All right. Thanks, Mr. Rogers. Thanks, panel. 1012 Thank you, Mr. Chairman, Mr. Birchenough. Those are my questions. 1013 MR. SOMMERVILLE: Thank you, Mr. Brett. I note that Mr. Brown and Mr. Vegh are present. I think, Mr. Brown, you have -- as I understand it -- 1014 MR. BROWN: A request. 1015 MR. SOMMERVILLE: A request for an extension of time. 1016 MR. BROWN: That's correct. Good afternoon, Mr. President -- Mr. Chairman, Mr. Birchenough. 1017 Mr. Penny, good afternoon to you. 1018 MOTION BY CORAL ENERGY: 1019 SUBMISSIONS BY MR. BROWN: 1020 MR. BROWN: Pursuant to Coral Energy's application before you a week and a half ago, as a result of that, you allowed the late intervention and directed that Coral Energy filed its evidence by today. I wish to advise the Board and other participants in the proceeding, we've been working very, very hard on the evidence. It is not ready. It will be ready Tuesday and will be ready to be filed Tuesday. So I'm here before you this afternoon to ask for leave of the Board to extend the time for filing until Tuesday. 1021 MR. SOMMERVILLE: Thank you. 1022 Mr. Penny. 1023 SUBMISSIONS BY MR. PENNY: 1024 MR. PENNY: Well, if Mr. Brown's client can't do it, they can't do it. You'll be aware that our phase-2 witnesses are the people that have to deal with this, Mr. Kitchen and his group, and I think you're also aware that, among many other things that Mr. Kitchen has to do in the context of this case, it takes about six weeks to do a blue-page update from the phase-1 blue-page update, and we've advised you earlier that it's our intention to do that. It will be available during the course of this case, and obviously in advance of the testimony of the phase-2 panel. 1025 So they have a lot to do. And if -- I'm not sure whether Tuesday morning is being represented as the date for delivery because that's the first business day or not, but if I could just ask Mr. Brown, through you, that if that evidence is available Saturday or Sunday, that we get it then, because I know for a fact that Mr. Kitchen is working on the weekend, and the sooner he gets this, the better. 1026 MR. BROWN: Mr. Chair, I appreciate Mr. Penny's situation, and certainly I know that people are working hard on this. We will be working hard on the weekend as well to finalize it. So Tuesday will be the first opportunity where the material will be finalized and ready to be filed. We've set out a schedule for work over the weekend that will allow us to file on Tuesday. 1027 MR. SOMMERVILLE: Mr. Penny, is an extension of time with respect to Union's response to this evidence, is that something that may be of interest to you? 1028 MR. PENNY: I'm not sure it will help, because it really just leads to a sort of domino effect. 1029 MR. SOMMERVILLE: If that does arise, feel free to raise that with the Board. 1030 MR. PENNY: Yes. 1031 RULING: 1032 MR. SOMMERVILLE: The Board has given this -- because we had some knowledge of this request prior to Mr. Brown's actual presentation of it these last few minutes, the Board will grant the request. We're mindful of the competing interests that we have here, and we certainly expect Tuesday to be the day in which this is presented. And, Mr. Penny, we would -- if this causes some additional difficulties for your people, we'd be certainly interested in hearing about that and we'd be sympathetic to that. 1033 MR. PENNY: Thank you, Mr. Chairman. We'll advise the Board if any further adjustment is necessary. 1034 MR. SOMMERVILLE: Mr. Vegh, I don't think this ruling affects you in any way. 1035 MR. VEGH: With respect, sir, it does. The original timetable would have been that Coral would have filed today. As a practical matter, the only opportunity, then, for evidence for my client, since their schedule was for Tuesday, would have been to react to Coral's evidence. Now, if this had arisen in a typical case and typical time frame, there's no doubt that Coral -- sorry, that TransAlta would have had the opportunity to react to a rate proposal that has them in their cross-hairs. And if the applicant had brought that rate proposal, they would react in the normal course; if another intervenor had brought it, then they'd have to apply for leave, but I'm certain that the Board would have granted it, given the direct impact on the customer. 1036 So the situation now is that if -- and TransAlta supports Coral's application. If they need more time, that's the most important thing. It's better to have better evidence on this than to just rush it through. But if Coral is not going to file until Tuesday, then if TransAlta's deadline for filing some evidence is not extended, then as a practical matter TransAlta would have no opportunity to address this issue in evidence. 1037 My request would be that TransAlta be granted an extension of time to file evidence in reaction to this to the end of the week, to Friday, and then Union's response could deal with both Coral's and TransAlta's position, if any. 1038 MR. SOMMERVILLE: Mr. Penny. 1039 MR. PENNY: I have no submission on that, Mr. Chairman. 1040 MR. SOMMERVILLE: We'll grant that request, Mr. Vegh. 1041 MR. VEGH: Thank you. 1042 MR. SOMMERVILLE: The extension with respect to the TransAlta evidence. We do have a sense of slippage of dates, and that's a matter of some concern to us as we go forward. But I understand the request that you're making and we'll proceed on the basis that we've indicated. 1043 Once again, Mr. Penny, if Mr. Vegh's request occasions any specific difficulties, please advise the Board accordingly. 1044 MR. PENNY: Yes. 1045 MR. SOMMERVILLE: Our interest is, as Mr. Vegh indicated, to provide the right opportunity for the issues to be considered, and that may require some further rulings as we go down the road here. So we'll deal with those as they may or may not arise. 1046 MR. PENNY: Yes. We'll review the situation with Mr. Kitchen and his group and advise the Board if we require any further adjustment. 1047 MR. SOMMERVILLE: Are there any further submissions arising from our ruling? 1048 MR. PENNY: Not from me. 1049 MR. SOMMERVILLE: Thanks a lot. 1050 MR. BROWN: Thanks very much, Mr. Chair. 1051 MR. VEGH: Thank you. 1052 MR. SOMMERVILLE: We'll take 15 minutes. We'll reconvene at 3:00 to proceed, I think, with Mr. Shepherd, Ms. Singh, Mr. Dingwall, Mr. Scully. 1053 MR. SCULLY: I have no cross-examination. 1054 MR. SOMMERVILLE: Mr. Rowe? 1055 MR. ROWE: No, sir. 1056 MR. SOMMERVILLE: And Mr. Moran. 1057 MR. MORAN: That's right. 1058 MR. SOMMERVILLE: So we have 3:00. We'll reconvene and see where we get. Thank you very much. 1059 --- Recess taken at 2:45 p.m. 1060 --- On resuming at 3:03 p.m. 1061 MR. SOMMERVILLE: Please be seated. Thank you. 1062 MR. PENNY: Mr. Chairman, before we begin, I think we're in a position to answer on the transcript Mr. Aiken's Interrogatory N.5.3 that had to do with the evidence of the Wholesale Gas Service Purchasers Group and their monthly volumes on the table on page 1 of their evidence. 1063 MR. SOMMERVILLE: All right. 1064 MR. ROGERS: Yes. I've reviewed this over the break and I'm prepared to accept it into the evidence. 1065 MR. SOMMERVILLE: Thank you, Mr. Rogers. 1066 MR. PENNY: That will be reflected with the phase 2, blue-page update that Mr. Kitchen is preparing. 1067 MR. SOMMERVILLE: Mr. Shepherd. 1068 MR. SHEPHERD: Mr. Chairman, one of the advantages of being late in the batting order is that you don't have much left to do. I only have two questions. 1069 UNION GAS LIMITED - PANEL 4; RESUMED; GARDINER, POREDOS, ROGERS 1070 P.GARDINER; Previously sworn. 1071 S.POREDOS; Previously sworn. 1072 B.ROGERS; Previously sworn. 1073 CROSS-EXAMINATION BY MR. SHEPHERD: 1074 MR. SHEPHERD: They're both for Mr. Gardiner, I believe. 1075 Mr. Gardiner, and I say this with the greatest of hesitation, could you please bring up Exhibit J.7.1. 1076 MR. GARDINER: I have it. 1077 MR. SHEPHERD: Now, there have been a number of discussions about the numbers on this piece of evidence, and I listened to them yesterday and the day before, and I read the transcripts last night and the night before and I still don't understand, so I'm going to ask again. 1078 That $104 million number, that's the delivery-related deficiency that you currently project for the test year; right? 1079 MR. GARDINER: This is Pat Elliott's interrogatory and response. Unless I can get some assistance from someone else -- let me read the question. 1080 It is the delivery-related deficiency of $104 million. 1081 MR. SHEPHERD: Thank you. This says that the amount of that deficiency that's driven by the change in the weather methodology is $27 million. Now, do I understand correctly that, if this Board decides not to allow the company's request to change methodology, that the deficiency will go down by $27 million? 1082 MR. PENNY: Mr. Chairman, we have been over this. Mr. Shepherd may not -- I mean, we have been over this. It's, I think, been said by me and by the witnesses on several occasions that the actual revenue deficiency that translates into 2004 -- this is the difference between '99 and 2004. The 2004 stand-alone methodology only, weather methodology only number is not actually in this answer. It's $20.4 million. 1083 MR. SHEPHERD: Mr. Gardiner, do you adopt that as your evidence? 1084 MR. GARDINER: Yes, I do. I already have. 1085 MR. PENNY: He already has. It's in his evidence. 1086 MR. SHEPHERD: Well, okay. Then I don't understand because I'm reading an answer that says the delivery-related deficiency is $104 million, and there's four parts to it, and one of the parts is $27 million -- 1087 MR. PENNY: Yes, but you have to read the opening line. It says: "Following key items when compared to the 1999 forecast underlying current rates." 1088 MR. SHEPHERD: Well, sorry, here's my difficulty: There's $104 million deficiency. Is there any doubt about that right now? That's the number, isn't it, in the blue page? 1089 MR. GARDINER: I can only say that that's the number that's on the interrogatory response as prepared by Pat Elliott. 1090 MR. SHEPHERD: And so what I'm trying to figure out is, if the Board says, no, you can't change weather methodology, does that mean that the ratepayers get to keep $20.4 million or $27 million? And as I understand this, it says they get to keep $27 million. And if I'm wrong, I'd like to understand why. 1091 MR. GARDINER: I'm sorry, you'll have to ask Pat Elliott that question. 1092 MR. SHEPHERD: Okay. 1093 My second question is -- and I wonder if you could pull up the yellow pages, Exhibit A, tab 11, schedule 1, page 4. 1094 MR. GARDINER: I have the information in front of me. 1095 MR. SHEPHERD: Thank you. So there's a number there, $885,171,000; do you see that? 1096 MR. GARDINER: On line 7, yes. 1097 MR. SHEPHERD: Okay. And that number is calculated, I assume, assuming that your rates are set based on 20-year trend weather methodology and, in fact, the degree days for next year are the number that that methodology predicts; correct? 1098 MR. GARDINER: The gas sales revenue would be based on the tool for normal from the 20-year declining trend, and this -- since it's the yellow page and I've not seen this before now and it's prepared by someone in the finance department, I don't know which rate schedules they've used. This is a schedule that would also be best asked to Pat Elliott. 1099 As I mentioned, the volumes that are in the gas sales would be from the demand forecast which would have 2-04 normal in it. The rate schedule that was used and the cost of gas that was used, you'd have to ask the people that prepared this schedule which set they used. I'm not aware of that. 1100 MR. SHEPHERD: Well, I'm actually -- what I'm going for is actually quite a different question, and that is, I'm trying to understand what actually happens if the 20-year declining trend method is used, as you proposed in your application, and it turns out that actual degree days in the test year are the number that the 30-year average would predict, which is, what, about 3,950 or 4,000, something like that? 1101 MR. GARDINER: That is the $20.4 million that Mr. Penny mentioned. 1102 MR. SHEPHERD: Well, no, that's not the question I'm asking. So here's the question: Can you please undertake to provide us with the number equivalent to that 885,171, the number that would be the case if your application is approved as is and the degree days are the number predicted by 30-year average. I'm asking for an undertaking. 1103 MR. GARDINER: Mr. Chairman, can I accept on behalf of another panel that's coming later? I don't know if I can do that. 1104 MR. SOMMERVILLE: Well, you can accept on behalf of the applicant. 1105 Mr. Penny, do you -- 1106 MR. PENNY: Could I just have a moment? 1107 MR. SOMMERVILLE: Sure. 1108 MR. PENNY: As I understand it, the difficulty is it won't change just one line on the pro forma, it will change the transport volumes, the distribution volumes, and the cost of gas supply, so it flows through everything. So it's a big job to redo these pro formas. You can't just plug one thing in and leave the rest alone. 1109 MR. SOMMERVILLE: It is an analysis that I think would be of interest, and I wonder if, given Mr. Gardiner's comment, it might not be appropriate for this to actually be an undertaking for a subsequent panel -- or for the company to be executed by its subsequent panel. 1110 MR. PENNY: Perhaps the best thing to do is this, can we leave it on this basis: I'd like to get a clear statement from Mr. Shepherd of what he wants, and then why don't we just take that away and think about it -- it's not Mr. Gardiner's bailiwick in any event -- and we'll see -- we'll see what we think we can do with that, and then come back and have a discussion with Mr. Shepherd about it and advise the Board. 1111 MR. SHEPHERD: I can't tell from what Mr. Penny has just said whether he's willing to give the undertaking or not. 1112 MR. PENNY: Let us know what it is you want, and let us go and consider whether it can be done, and then we'll deal with it in due course. 1113 MR. SOMMERVILLE: If the results of that exchange doesn't leave you where you want to be, Mr. Shepherd, you can raise that subject with me and Mr. Birchenough again. 1114 MR. SHEPHERD: Thank you. Mr. Chairman, the request, then, is that the company recalculate that page to identify what the number 885,171 would be, assuming two things: First, that their application is assumed as filed, as currently filed; and secondly, that the actual degree days in the test year are the number of degree days projected by the 30-year average method. 1115 MR. SOMMERVILLE: Thank you. 1116 MR. SHEPHERD: Mr. Chairman, those are my questions. With your permission, I'll take my leave. 1117 MR. SOMMERVILLE: It seems a little unfair, Mr. Shepherd, that you get to leave. 1118 MR. SHEPHERD: I'll listen over the Internet, I promise. 1119 MR. SOMMERVILLE: Thank you. 1120 Ms. Singh. 1121 CROSS-EXAMINATION BY MS. SINGH: 1122 MS. SINGH: Good afternoon, counsel. I'm counsel for the Canadian Manufacturers and Exporters, and I have a very few questions with respect to the impacts using the 20-year trend on capital asset planning decisions by Union. 1123 I'm just going to refer to the transcript of October the 7th, at paragraph 942. 1124 When panel number 3 was testifying, and I'll read it. If you want to turn to it, I'm happy to wait. The October 7th transcript, and it's paragraph 942. It's Mr. Shepherd's cross-examination of Mr. Fogwill. Mr. Shepherd asked: 1125 "I wonder, Mr. Fogwill, if you can turn to page 5 of our cross-examination materials. This is Exhibit M.2.3. All this does, and this is just math, it just extends your current trend line that's predicting 2004 out another 20 years. The trend that you've identified in the last 20 years implies that 20 years from now will be at around 3,200 heating degree days. Is that what you think is going to be the case?" 1126 And Mr. Fogwill responded: 1127 "No, no, this is inaccurate in terms of representing our method. Our method is recalculated each year, so the slope of this line is only good for one year, our test year, and then next year we will recalculate it again and we'll get a new slope and it will be good for the following year, and so on and so forth." 1128 Mr. Shepherd: 1129 "Didn't you just tell us that it had no predictive value for next year?" 1130 And Mr. Fogwill responded: 1131 "We're not identifying it as being used as predictive value. We're identifying it as being used as a benchmark for planning purposes. And what that benchmark will say is that if you recalculate the 20-year trend, you're going to get a new slope of the line and therefore a new estimate." 1132 My question for the panel is, given this exchange, could you explain how Union goes about making its capital asset planning decisions, including storage and pipeline expansion decisions, on the basis of identifying the slope of the line for one year point on a 20-year trend line and not extrapolating out beyond that one point? In other words, just for clarification, how does Union plan when they know the slope of the line will be different in the next year? 1133 MR. PENNY: That, Mr. Chairman, is a question -- if it's planning with respect to capital assets, that is Larry Hyatt's responsibility, he's on panel 9, and if it's planning with respect to gas supply matters, that's the gas supply pannel, Mr. Isherwood, Simpson, and Newbury. 1134 MS. SINGH: Thank you, Mr. Penny. 1135 I'm wondering if this is your evidence for this panel, at C.1, tab 4, page 7, you state that: 1136 "The company has been purchasing --" 1137 MR. ROGERS: I think we're still looking for it here. 1138 MS. SINGH: Sorry, it's actually page 7. 1139 MR. MORAN: Is that 7 of 31 or 7 of 7? 1140 MS. SINGH: Page 7 of 7, the white pages. Maybe I'll go back to page 6, to put it in context for you. 1141 MR. ROGERS: We're trying to figure out whose evidence it is. 1142 MS. SINGH: It's Mr. Gardiner's evidence. 1143 MR. PENNY: Sorry, which page now? 1144 MS. SINGH: It's page 6, underneath the heading "Impacts of Using the 20-year Trend." 1145 MR. GARDINER: Yes, I have the evidence. 1146 MS. SINGH: In your evidence, you state: 1147 "The weather normalization assumption affects forecasted customer consumption, storage, and transportation allocations, gas supply planning, and operations. Using a 20-year trend appropriately reduces the forecasted volumes to a level more in keeping with actual experience, and it would increase rates in comparison to using the 30-year simple average." 1148 Again, you go down to the next paragraph and you say: 1149 "The definition of normal weather, and therefore its estimation, is a significant factor in the design and development of long- and short-term operational plans. It provides support for the upstream capacity required to serve the entire Union system. 1150 "Union has been optimizing the upstream transportation requirements using the 20-year trend. The result is that the company has been purchasing or contracting for fewer assets than would be required if Union continued to use the 30-year average method. This will result in an overall cost reduction to infranchise customers both in terms of the expense of that capacity and through the avoidance of unabsorbed demand charges." 1151 My question for this panel, then, or Mr. Gardiner, I guess, because this is Mr. Gardiner's evidence, is would you agree with me that that last sentence would not be the case in the event that that 20-year trend underpredicted the temperature for any particular year? 1152 MR. GARDINER: I'm sorry, could you repeat the question? 1153 MS. SINGH: Would you agree with me that the proposition that this will result in an overall cost reduction and so on for infranchise customers would not be true if the 20-year trend method that you're proposing, in fact, underpredicted the temperature or heating degree days for any particular year? 1154 MR. GARDINER: This has been discussed earlier. But to summarize, because of the symmetry inherent in the 20-year declining trend, there will be some years where the weather will be colder than the trend and, yes, the revenues would be higher. There also will be years where there is the inverse. So the future will tell on what side of the line it results. So, yes, there could be, as the lady says, years where you have additional revenues and years where you have costs that aren't recovered. 1155 MS. SINGH: Thank you. I'm just concerned, Mr. Gardiner, that for the purposes of this Board setting its rates based on the 20-year trend, if Union is tying its capital asset purchasing decisions to the 20-year trend methodology such that, in your evidence, you state that the company would be purchasing or contracting for fewer assets, and this turns out to be wrong and every year you have to go back and revisit that decision, how prudent that decision is in terms of the effects for ratepayers. 1156 MR. GARDINER: This, again, was discussed. The issue is the symmetry. And I refer you to the thermometer figure, figure 1, on Exhibit C.1, tab 4, page 3 of 17. We've seen it before, Mr. Chairman. As we discussed yesterday, the historic analysis indicates that we have more symmetry with the 20-year trend than we do with the 30-year average. We heard the scientists on the first panel talk about the continuation of a long-run warming trend. And we believe that the 20-year trend is the best method to provide that symmetry. So this has been discussed. 1157 MS. SINGH: And is it your evidence, Mr. Gardiner, that symmetry is the correct objective for capital asset planning purposes for Union? 1158 MR. GARDINER: To respond to your question, the symmetry involves the general service demand forecast which is part of the demand. There are other demands. And you're asking the transmission supply planning people -- I think I've misnamed them, the transmission planners, about how they plan. I can't really address that. The general service demand forecast is a component of the total demand, and we need symmetry. But I really can't answer how symmetry goes into their plans. It goes into my demand forecast. 1159 MS. SINGH: Well, thank you, Mr. Gardiner, I guess I'll come back and revisit that question with, I think, panel number 9, Mr. Penny indicated. 1160 I guess I'm just concerned, Mr. Gardiner, and this is what my evidence is about, is that if the 20-year trend method is very variable, particularly from a regional perspective, according to the evidence that was given by Dr. Weaver, and the asset-planning decisions of the company are tied to that 20-year trend and they're inaccurate for a particular year, that the ratepayer would need to go out into the open marketplace to buy the assets that were no longer planned for any particular year for them and that could be an expensive proposition. 1161 MR. PENNY: Well, this is not the time for argument. It's equally true that if the company, on the 30-year average, acquires too many assets, then there are unabsorbed demand charges which are then visited upon the customers. So, I mean, I think we've established that Mr. Gardiner is not the appropriate person to speak to design planning for the transmission system. He can speak to the general service forecast. So, in my submission, there's no question in that last statement, in any event, so why don't we move on. 1162 MS. SINGH: Okay. Thank you, Mr. Penny. Thank you, panel. Those are my questions. 1163 MR. SOMMERVILLE: Thank you, Ms. Singh. Ms. Singh, you too can abandon ship at any point. 1164 MS. SINGH: Thank you, Mr. Chair. 1165 MR. SOMMERVILLE: Mr. Dingwall. 1166 CROSS-EXAMINATION BY MR. DINGWALL: 1167 MR. DINGWALL: Thank you, Mr. Chairman. 1168 Good afternoon, panel. My name is Brian Dingwall. I'm counsel on behalf of Energy Probe today, and hopefully my introduction will take as long as my questions. 1169 I'm going to limit my questions to Mr. Gardiner and Mr. Poredos. 1170 Starting with you, Mr. Gardiner, there's been a significant amount of discussion over the last couple of days which relates to variable factors that affect the projected consumption for the M2 and general service rates, and I believe -- and I'm paraphrasing -- that these variable factors include price sensitivity of residential customers, demand-side management, marketing on behalf of the company, and energy efficiency. 1171 Are those basically the four food groups here? 1172 MR. GARDINER: You have the food groups correct. 1173 MR. DINGWALL: Great. Now, with respect to demand-side management, my understanding is that on page 2 of 34.33, of one of the interrogatories -- I'm trying to remember the intervenor's name, but that's okay, it's late in the day -- that there's a display of what the demand-side management affect on the general service forecasts are; is that correct? 1174 MR. PENNY: Could we have that reference again, Mr. Dingwall. 1175 MR. DINGWALL: I believe it would be J.7.34 -- or is it J.34.34? 1176 MR. SOMMERVILLE: 34.33? 1177 MR. POREDOS: Could you repeat the question so we could relate it to 34.34, make sure it's the right one? 1178 MR. DINGWALL: Sure. Is the projected effect of DSM reflected in the projected normalization for 2002, 2003, and 2004, on page 2 of that exhibit? 1179 MR. GARDINER: Yes, it is. And that was the table we were looking at earlier. And as we discussed yesterday and today, you can see the DSM impacts -- the DSM impacts in each of the years. 1180 MR. DINGWALL: And that's reflected in the first column on that page that says DSM? 1181 MR. GARDINER: That's correct. 1182 MR. DINGWALL: And the marketing plan is the company's efforts to market natural gas to existing customers; is that correct? 1183 MR. GARDINER: To existing and new; to increase use of natural gas clothes dryers, stoves, fireplaces, and other gas appliances that are used in the commercial and residential and industrial markets, small markets. 1184 MR. DINGWALL: And is the other matter that we discussed before, retail price sensitivity, reflected in the reasonability calculations contained in the subsequent columns? 1185 MR. GARDINER: The efficiency and the price variables and the weather are contained in the econometric estimates which are under the NAC columns under equation 1 and 2, and then those are averaged to get the NAC estimate that's shown under base case NAC. And then the net sum of the DSM marketing plan and water heater is then added to that econometric estimate, which is the average of the two, to come up with the final NAC, 2,642. 1186 MR. DINGWALL: Now, where in that calculation is price sensitivity reflected? 1187 MR. GARDINER: It's equation 1 and equation 2, because in both equations there's a retail energy price. 1188 MR. DINGWALL: Now, I believe there's been some mention that when prices change, there's a degree of sensitivity that customers exhibit either based to consumption or based around a customer's decision to leave system supply; is that correct? 1189 MR. GARDINER: The econometric equations assume a -- are based on a burner tip or price at the household that include a March QRAM, which does not change over time, and reflect the rates when the -- the Union Gas delivery rates at the time the forecast was prepared, which was in that March to May period. So the forecast assumes a constant price over time. And as I mentioned this morning, the reason we did not put a price forecast in was we felt that the variation around the burner tip was not significant enough, given the pricing elasticity of demand for those markets, and also -- and given that the QRAM at the time was unknown, what the October QRAM or July QRAM. We had the March one. So we said, Okay, let's fix the price assumption. 1190 MR. DINGWALL: So if we can translate that from statistic language into plain speak, does that mean that you believe customers do not exhibit behaviour that would either move them away from system gas to direct purchase or reduce consumption in the event of price changes? 1191 MR. GARDINER: Okay, two things: All the demand forecast we're talking about here is the combined market. These are the usage of all customers, be they system supply or direct purchase. That's the first thing. There's no distinction in these numbers. 1192 Yes, customers are price-sensitive, but in the forecast, our forecast assumption is the price is fixed. A 10 percent change in price at the burner tip, is about a .7 percent change in demand for a residential customer, which is a small amount on an annual basis. 1193 MR. DINGWALL: Moving on to Mr. Poredos. Your evidence had to do with the forecasted revenue associated with transactional services; correct? 1194 MR. POREDOS: That's correct. 1195 MR. DINGWALL: With respect to next year's forecast, do any elements of that forecast take into account any transactions which would be multiyear in nature? 1196 MR. POREDOS: There would be -- there would be transactions, and if they are multiyear, they would be reflected both in 2004 and beyond. We have long-term storage contracts which go beyond 2004 which are reflected in the long-term premium. We have M12 transportation contracts that are longer-term contracts that would, in fact, have revenue beyond 2004. So those would be reflected in the future in those accounts and recovered through the deferral mechanism. 1197 MR. DINGWALL: And for short-term storage contracts, I take it they would include amounts where they would be attributed partly to year 2003 and partly to year 2004; is that correct? 1198 MR. POREDOS: Most storage contracts will start on April 1 of a year, because we're at the lowest position in our storage, and will extend to April 1 of the next year, so, yes, they will cross a physical -- a calendar year, let me put it that way. Those would be reflected in that way. There are some storage contracts that may be very short term where they might buy the storage in, say, July and have a contract through April, so they also would be extended across if it's a one-year contract. So it would be between April and April. 1199 MR. DINGWALL: At this point in time, has the company completed contracting for the 2004 winter season? 1200 MR. POREDOS: Could you clarify contracting what for the winter season? 1201 MR. DINGWALL: Storage. 1202 MR. POREDOS: You mean our storage that we'd be selling into the market? 1203 MR. DINGWALL: Yes, your storage you would be selling into the market. 1204 MR. POREDOS: The majority of the storage would be thus far sold and customers are filling it as we speak because they're getting ready for the winter. So the majority of storage thus far would be on the books. There may be some storage late season that we would release over the next two or three weeks that would be a very short-term injection period for the customer that would then also go into the winter. 1205 MR. DINGWALL: And would that likely be the end of when the company would do its principal contracting for storage for the test year? 1206 MR. POREDOS: That's not necessarily so. We would be selling storage beginning with contracts next April, which would then extend, within 2004, some revenues. And we've reflected that in those accounts. 1207 MR. DINGWALL: Does the company anticipate filing any update with respect to its forecast for 2004 prior to the end of the evidentiary phase of this hearing? 1208 MR. POREDOS: I'm not aware of a requirement to submit another forecast. 1209 MR. DINGWALL: Well, would it be fair to say that by the first week of November, you would know, with some higher degree of certainty, what you contracted for next year? 1210 MR. POREDOS: We would be much closer to the end state. There would be off-peak storage and other services that may still be offered and sold during those last two months. But the majority of the year, yes, would be done and over with. 1211 MR. DINGWALL: And I take it that because of the planning importance of these transactions, that they'd be reported to your department on a fairly regular basis, where you'd have a significantly close-to-realtime knowledge of the transactions; would that be correct? 1212 MR. POREDOS: Yes. My department, in fact, is responsible for releasing the asset to the salespeople and then they go out and sell it. Now, if I release it, they may not have an opportunity to sell it. It will be on the shelf. But if the opportunity does show up, they will put it into the market. 1213 MR. DINGWALL: So if we were to take November 1, being a date at which this hearing is still ongoing, and on which evidence still presented, what kinds of reports or information would you have available at that point in time? 1214 MR. POREDOS: We would have our monthly update, internal information that we do monthly to manage the business. 1215 MR. DINGWALL: Which would, agree with me or not, sir, give you a better handle on the projected revenues associated with your transactional services for the test year and an indication of the additional capacity that may have been released into the market at that time? 1216 MR. POREDOS: We did the blue-page update just probably a month, maybe a month and a half, two months ago, which we had very up-to-date information to that point. There is not a lot of incremental business at this point that would be added to it. So our view of 2004 is pretty good. 1217 MR. ROGERS: Was the blue-page update, Mr. Poredos, just to clarify, was it five and seven, so up to May. 1218 MR. POREDOS: From a financial standpoint. But the forecast forward, we would have looked at it -- we would have had any information updated to look forward to say that we've updated the blue page. We did the update on blue page. 1219 MR. DINGWALL: Well, what we've just heard is that there's two different things we might be talking about. Let's try to be clear. The blue-page update was on the basis of financial information and forecasting; is that correct? 1220 MR. POREDOS: The financials that would have been recorded or the forecast would have been done on a five and seven at the time of blue page, that's correct. 1221 MR. DINGWALL: When you say five and seven -- 1222 MR. POREDOS: Five months of actuals. For 2003, we're talking about, you'd have five months of actuals at that point and seven months of forecast for the remainder of the year. 2004 is all forecast, other than long-term contracts that may be signed which is a normal revenue that would be reflected every year within those accounts. 1223 The account for 2003 will also be reviewed at the end of the year from a standpoint of the actuals, regardless -- into the new year, we will have to defer those, or split is the 75/25. 1224 MR. DINGWALL: So what's your timing to close the book on actuals from the time that you finish a quarter? 1225 MR. POREDOS: Finish a quarter? 1226 MR. DINGWALL: Or finish whatever relevant reporting period. You report to the ERO on a monthly basis, don't you? 1227 MR. POREDOS: I believe it was quarterly, but I can't answer that because I'm not the one who presents that information. 1228 MR. DINGWALL: From your perspective, sir, you'd be able to comment on what contracts were in existence and what capacity had been released and what transactions had come onto the books for your department; would that be correct? 1229 MR. POREDOS: In fact, we provided that as an interrogatory, which I believe was a five and seven forecast during IRs. 1230 MR. DINGWALL: Right, which would be five months of actual and seven months of forecast. We're now in October. I'm just wondering what additional information might be available at this time. 1231 MR. POREDOS: Well, we would have more up-to-date information if we were to forecast again. But that does not necessarily suggest that the 2004 forecast would change, because the only thing that's changing is 2003 actuals. 1232 MR. DINGWALL: So is it your evidence that the current status of the 2003 actuals do not, in your mind, give rise to any need to update or review the 2004 forecast? 1233 MR. POREDOS: When we looked at the blue page, an update of the five and seven, we were comfortable that that position is a reasonable forecast or a reasonable position for 2004. And that's what we've submitted. The 2003 final information will be submitted when the deferral accounts are disposed of. 1234 MR. DINGWALL: Would the information about what actuals you had be best addressed with the accounting panel? Let me just clarify that. 1235 MR. POREDOS: I suspect the accounting panel may have more information on that than I do, from a standpoint of the information they have. It's not going to change my forecast, though, from the standpoint of what we've done or what we've put forward for 2004. We feel comfortable with the forecast for 2004. 1236 MR. DINGWALL: Has the transactional services group given any -- entered into any multiyear transactions this year that have an impact on 2004, such as gas loans? 1237 MR. POREDOS: I don't believe we have any long-term gas loans that go into 2004, no. We would only be doing short-term gas loans. And, in fact, the majority of the gas loans that we did forecast have not been done because of the credit issues. 1238 MR. DINGWALL: What's the normal, just so I understand, term of a gas loan? 1239 MR. POREDOS: A gas loan can be inter-season or one year to the next, summer to summer, winter to winter. The majority of our gas loans, under normal years, would be shifting gas from one month to the next, or shifting a winter inventory in November that I don't require in that month to a month that I might require it, and then getting the differential for that. From November to December is an option. 1240 MR. DINGWALL: Okay. So when we're talking about gas loans, these are not something that the company would enter into from this year, say, a year out or two years out. 1241 MR. POREDOS: Not under normal circumstances, no. We've used that option when we did mitigation for warmer-than-normal weather two years ago. We no longer have those on the books. 1242 MR. DINGWALL: Okay. Now to move up to the 10,000-foot level for my final question. It seems that we've heard from a number of intervenors in this proceeding that there's a concern with how the company's forecasts and the lack of consensus with individual intervenors who might be large users and contract demand customers, how disagreement with respect to those forecasts seems to create business issues between the company and market participants that haven't been resolved and seem to be coming to the table. Why would the company not give full weight to estimated consumption numbers from contract customers? 1243 MR. ROGERS: As I have said prior to this, in terms of the large group of 76 industrial and power customers, they are all very large and sophisticated. We do sit down and talk with them as a group. We have those discussions. What we do is take their input and we look at any significant changes to their operation, up or down, and then we apply that discussion to their historical trends. We look at the CD levels that they have and whether they're going to change because of any of that significant change. And that's what we incorporate in the forecasts. And if we take out the issue of the power markets, which have been much more difficult to measure, and you remove the alternate fuel swing-loads that we've talked about prior and put on the record already, we just come in very closely, then, to the forecast. We don't have -- could you give me an example of someone who has reached a point where they don't want to reconcile the contracts? We've had discussions and we reconcile with the customers. So I'm not sure what the -- there seems to be inference or seems to be innuendo. What I don't have is a specific example of someone who, you know, we haven't reached an agreement with on this. So I'm not sure how to answer it any farther than that. 1244 MR. DINGWALL: I guess, in looking at the continual appearance of Kitchener, I'm not sure about that. But I guess the larger question, moving from that, is has the company thought about implementing or discussing some form of accelerated dispute-resolution procedure to take away some of the uncertainties that are associated with having an obligation to negotiate within contract customer classes? 1245 MR. PENNY: Well, sorry, but there's about three different things going on there. There is no obligation to negotiate the demand forecast. The question is kind of meaningless, Mr. Chairman, in my respectful submission. 1246 MR. SOMMERVILLE: I think the question is appropriate. Mr. Dingwall, if you could just rephrase the question and Mr. Gardiner, I'm sure, can provide you with an answer. 1247 MR. ROGERS: I think it's Mr. Rogers. 1248 MR. SOMMERVILLE: Mr. Rogers, I beg your pardon. 1249 MR. DINGWALL: Okay, I'll take out the preamble, but let's try this: Has the company considered, or is the company likely to consider adjusting the language of its contract class customer contracts to implement an accelerated form of dispute resolution? 1250 MR. ROGERS: Let me answer it this way: The new balancing proposal that is being put forward and will be discussed in this case but will not come into service, as I understand it, the request, until early 2005, will require the customer and Union to agree on forecasts because we will require more than just the annual once-a-year balancing service for final customers that we have now. 1251 So while we are changing our balancing provisions, that will change the need for a more rigorous and tighter forecast from agreement between both the customer and Union. And so that answers the question of will we be coming to that. I believe that that is the future. In fact, we're going to work towards dry-running that operation this year to help customers understand how to become more rigorous and more focused and diligent, quite frankly, on providing that to us on, rather than just an annual basis, a more frequent basis. 1252 So the answer, I believe, to your question on that is that we're moving in that direction, and we will be fully there, if the Board approves the new methodology, by 2005. But up until now, with annual provisions to balance once a year, with plus or minus 4 percent tolerance, we have not had the -- customers haven't provided as much emphasis on this as they probably will, going forward. And so we have to use judgment of what their operations are like, what similar customers are like, across those large customer classes, and put that forward into the aggregated forecast to closely reflect what we need so that Mr. Poredos and his group, and others, can plan for the assets required to provide that service. 1253 So I believe that's -- that's how it works today and that's -- we will be moving to probably a tighter forecast of demand and supply as we go forward. 1254 MR. DINGWALL: Has the language for that contract been developed yet, Mr. Rogers? 1255 MR. ROGERS: The language for the contract would be the -- that would be covered by Mrs. VanDerPaelt in her -- when she appears before the Board, so that I would defer to her for that review. 1256 MR. DINGWALL: Thank you very much, gentlemen. Those are my questions. 1257 MR. SOMMERVILLE: Mr. Dingwall. 1258 Mr. Moran, I think you have the distinction of being the last one up. 1259 MR. MORAN: Thank you, Mr. Chair. 1260 MR. DINGWALL: As a preliminary matter, I would like to follow the other rats off the sinking ship. 1261 MR. PENNY: The truth comes out. 1262 MR. SOMMERVILLE: Your characterization, not mine, Mr. Dingwall. Feel free to excuse yourself at any point in time. 1263 MR. DINGWALL: Thank you, sir. 1264 MR. MORAN: Now that he's backed me up against the wall for my cross-exam. 1265 CROSS-EXAMINATION BY MR. MORAN: 1266 MR. MORAN: Panel, my questions and your answers are the only thing between us and the long weekend, so with that shared responsibility, I'd like to start with you, Mr. Gardiner. 1267 First, I want to spend a couple of minutes, not very long, just a couple of very basic -- well, maybe not so basic, but some issues of statistics, given that you work on a daily basis with statistics. These questions aren't intended to necessarily point at any criticism of the data that you're relying on or the method that you followed. I just want to get some general propositions out. 1268 The first question: There was a reference to heteroskedascity a little earlier. I take it you understand what that means? It has to do with increasing or decreasing variability in your data points; right? 1269 MR. GARDINER: Correct. 1270 MR. MORAN: And there's a way to test for that, obviously; right? 1271 MR. GARDINER: Yes, there is. 1272 MR. MORAN: And the problem that it creates, if you have that problem, is that if you want to use the data and do linear -- ordinary linear regressions, the outcome is questionable; right? 1273 MR. GARDINER: The outcome may be problematic. But my charts that we've reviewed this morning indicate that it's not a big problem. 1274 MR. MORAN: Remember, I indicated earlier on this isn't necessarily aimed at any specific charts or evidence you've been involved in. I want to stick with the general principles. So if significant heteroskedascity is observed in the data points, then you should look and see what it is and whether it should be taken out; right? There are ways to remove it. 1275 MR. GARDINER: Sometimes you have to live with it too. 1276 MR. MORAN: Right. And if you have to live with it, I guess that's one situation; but if there's a way to deal with it, then you would do that, wouldn't you? 1277 MR. GARDINER: If there's a way to deal with it by respecifying the equation, that's an avenue. 1278 MR. MORAN: Right. And if it continues to be in the data and isn't dealt with, the problem that it creates is that performing ordinary linear regressions leads to statistics that might not be valid; right? 1279 MR. GARDINER: If there is a... 1280 MR. MORAN: As a general proposition. 1281 MR. GARDINER: The textbooks will indicate that. 1282 MR. MORAN: Thank you. All right. And moving to a second concept, stationarity, you're familiar with that concept? 1283 MR. GARDINER: Yes, I am. 1284 MR. MORAN: If you have data that shows a trend, that's considered to be non-stationary; right? 1285 MR. GARDINER: I'm sorry? 1286 MR. MORAN: If you have data that shows a lot of variability -- if the mean isn't constant and the variances aren't considered the same, it's considered to be non-stationary; right? 1287 MR. GARDINER: That is a concept, yes. 1288 MR. MORAN: And the problem that that creates, if you have non-stationarity, running a regression using that data also gives questionable results; right? 1289 MR. GARDINER: It can create, yes. 1290 MR. MORAN: Right. And there's a way to test for it and correct it, right, if that's the case. 1291 MR. GARDINER: Yes. 1292 MR. MORAN: That's all for the statistics lessons for today, Mr. Gardiner. 1293 I now want to move very briefly to you, Mr. Poredos. Maybe I misheard you, I think you indicated that when it comes to gas loans, you don't have anything that's longer than a year, typically? 1294 MR. POREDOS: I don't believe that at this moment we have any loans outstanding that are longer than being within the season. 1295 MR. MORAN: All right. But previously you've had gas loans that are longer than a year. 1296 MR. POREDOS: And that was done to mitigate the length we had because of the warmer-than-normal winter where we had gas in storage and it did not get burned by the customers we had bought it for and we had to mitigate that position. 1297 MR. MORAN: Okay. So the rationale behind longer gas loans would be to mitigate a problem associated with warmer-than-usual winters? 1298 MR. POREDOS: And those gas loans did create value for the customers. 1299 MR. MORAN: All right. Could you turn up Exhibit J.1.14, the IR response to Board Staff. 1300 MR. POREDOS: Yes, I have that. 1301 MR. MORAN: All right. In this exhibit, in the answer, we see a list of gas loans that were made to Enron North America Inc., and they appear to be for longer than one year, except for the last one; right? Sorry, the second to the last one. Two of them are longer than a year -- 1302 MR. POREDOS: Well, the period that these cover, there are two years that were warmer than normal, in fact, that we had additional gas so, yes, they could extend longer than one year and be a two-year loan, correct. 1303 MR. MORAN: Right. Now Enron North America is an energy trader; right? 1304 MR. POREDOS: Correct, or was. 1305 MR. MORAN: No longer is. As a result of the Enron problem, you had to write-off some of these loans; right? 1306 MR. POREDOS: Yes. There was a loan failure on Enron. 1307 MR. MORAN: Okay. So to the extent that there might have been value for customers, that would have been cancelled out by the write-off? 1308 MR. POREDOS: In that particular one there may have been a loss taken on it. 1309 MR. MORAN: Right. And is that the $13 million figure? Is that the write-off? 1310 MR. PENNY: Sorry, what is the -- what does this have to do with the forecast is, I guess, my question, Mr. Chairman. This has to do with the deferral account disposition, it sounds like, of historic activity. 1311 MR. SOMMERVILLE: Mr. Moran. 1312 MR. MORAN: Mr. Chair, I'm prepared to move on. I was asking a more general question, given that Mr. Poredos is here and he's the only one that could speak to it, my question was basically coming up, and that was: What was the rationale behind making this kind of a loan to an energy trader? I appreciate it doesn't have anything specifically to do with this panel, but I don't have anyone else to ask that question of at this point. 1313 MR. SOMMERVILLE: It seems a reasonable question but one that perhaps should be asked of a subsequent panel that could speak to that specific kind of arrangement. 1314 MR. MORAN: I think that's the problem. I'm not sure that there is anybody coming up through the ranks that will be able to address this, other than Mr. Poredos, unless I can stand corrected. 1315 MR. PENNY: Mr. Poredos is on the gas supply panel and that's the context. 1316 MR. POREDOS: No, I'm not. 1317 MR. PENNY: Oh, sorry, I take it back. Anyway, I wasn't objecting to the question so much as posing the question of whether it had anything to do with the subject at hand. Maybe it will speed things up if we hear Mr. Poredos' answer. I think he's already answered it. 1318 MR. POREDOS: Can I hear the question one more time, please. 1319 MR. MORAN: In making a gas loan to a trader, I'm just trying to understand what the rationale was and what evaluation there was made of the risk associated with making a gas loan to a trader like Enron? 1320 MR. POREDOS: The decision to do any business with any trader or any company would be done exactly the same way, regardless of who the company is. The first issue is we would have to check the credit rating for that particular company and see how much credit they actually have with us or how much credit they've already used up. Under that -- under that scenario, then, if there's a transaction that needs to be done, we would either go out to bid or they would bid on that or offer us a price and we would do the transaction. It wasn't that we were doing it specifically with Enron or anyone else. They're just one of many traders at that point in time that had excellent credit ratings, which everybody would admit to at that point in time that were doing business with Enron. And subsequently we got caught in the same net that the rest of the energy business got caught in throughout North America with the Enron situation. 1321 MR. MORAN: Thank you, Mr. Poredos. 1322 Now moving over to your evidentiary neighbourhood, Mr. Rogers. I'm sorry. 1323 MR. ROGERS: In the neighbourhood. 1324 MR. MORAN: It's late in the day and I'm trying to keep everybody happy. 1325 MR. ROGERS: In the neighbourhood, I know. 1326 MR. MORAN: I think you've already talked about the issue of the nuclear plants coming on line and the relationship between that and gas-fired generation and what that has to do with your forecast. As I understand it, quite separate from that, as a result, I think, of a government initiative, there was an RFP for some stand-by generation which, I think, yielded some results, perhaps not as successfully as was hoped, but how was that taken into your forecast? 1327 MR. ROGERS: I believe there is an interrogatory that answers that. If I can beg your indulgence, I'll just have a quick look. I can speak to it without pulling it up. Basically, we had one -- we were successful in getting one contract customer that signed up. The period of time is very temporary; in fact, that contract is finished in April of 2004 and we don't expect that it will operate during the wintertime at all. It did operate during the blackout, which is good information, for a couple of days to help out when other things were down. But other than that, we're not expecting any operation from it at all. So there's nothing in the '04 forecast. 1328 MR. MORAN: Okay. So no impact with relation to your forecast at this point. 1329 MR. ROGERS: None. 1330 MR. MORAN: Speaking then that the nukes that are out of service and the plan to bring those back on, if I might just spend a minute with you on that. Could you just turn up Exhibit M.4.2. I think that's the easiest way to deal with this. 1331 MR. ROGERS: Yes, I have it now. 1332 MR. MORAN: And if we just look at the September 24, 2003 column on the first page of that exhibit, I think you were already asked questions about the Pickering A unit 3, which is now projected to come back on sometime in March of 2005, and I think you indicated that that didn't really make much of a difference overall to your end of the forecast. 1333 MR. ROGERS: That's correct. 1334 MR. MORAN: Looking at Pickering A unit 4, as of September 24th, there was an expectation that that would be fully available in September of this year. Am I correct in my understanding that that, in fact, hasn't happened? 1335 MR. ROGERS: I'm not completely up to speed, but I know that that unit was coming up and then had some difficulties. I'm led to believe they're temporary so that the target is still on track to come up and be on board by -- before the 1st of 2004. 1336 MR. MORAN: All right. One other question. I think you talked about OPG as a gas customer. Is Bruce Power a gas customer? 1337 MR. ROGERS: No, they are not, because we do not have a gas facility there. 1338 MR. MORAN: All right. 1339 MR. ROGERS: And on clarification, while we have Exhibit M.2 up, the April 2002 reference that we talked about yesterday, I did go back and do some homework, and while we used that directionally for our original forecast, we also used the 10-year outlook that was published in April 2002, and under all the scenarios, whether they be delayed availability of the nuclear units or the standard or the aggressive return of those units, there was still sufficient power in all cases, generation power, to meet the electricity loads without the large oil/gas generator being called on, so that that reconfirms, in my mind, our position and our forecast. 1340 MR. MORAN: All right. Now, I think you indicated in response to Mr. Rattray that in addition to relying on the IMO updates, you also rely on conversations that you have with the customer. I take it you discussed this with OPG, did you? 1341 MR. ROGERS: That's correct. And as we look back at -- if I look back at the information that we had from the customer, in each of the years '99, 2000, '01 and '02, their forecast of what they expected to burn was consistent with the minimum contractual commitment they had made, so that solidifies why we had our forecasts at that level for those years. 1342 Going forward, they're giving us information that, in fact, if we use the minimum and maximum ranges of that and divide it by two and get the average, and we apply the amount of oil-to-gas ratio that we've had even in the last year where we think that's probably aggressive for gas because gas prices are going up, even if we use that estimate, it correlates with our forecast of 65 10(6)m3. So we've come at it from at least two ways to correlate and solidify that number. 1343 MR. MORAN: All right. In the conversations that you've had with OPG, did they suggest to you that despite what they're saying to the IMO, we don't expect the nukes to be up and running for 2004? 1344 MR. ROGERS: They don't tell us that, because -- and I don't ask them. I mean, it's up to them when their units come back. I would reiterate, I would think the blackout would have kicked the wheels in motion to ensure that every possible thing that could be done would be done. 1345 The other thing that we haven't mentioned is that Darlington is down right now for a major overhaul and will be back up. They know it will be. It's just scheduled maintenance. That's going to increase the reliability of the Darlington nuclear which should, once again, make the whole nuclear fleet more reliable and push away the need to dispatch the oil/gas merchant plant. 1346 MR. MORAN: So OPG didn't tell you that things looked different from what the IMO -- 1347 MR. ROGERS: No, we weren't advised at all by our commercial contacts. 1348 MR. MORAN: All right. And did you have similar conversations with Bruce Power? 1349 MR. ROGERS: We didn't because they're not a gas customer. So they probably wouldn't have told us anyway. 1350 MR. MORAN: In terms of following up on the schedule that you seen in the IMO, you had no discussions -- 1351 MR. ROGERS: No, we don't have any relationship with the Bruce Power people. 1352 MR. MORAN: So for Bruce you rely solely on the IMO report? 1353 MR. ROGERS: Basically, yes. 1354 MR. MORAN: Thank you. All right. Just to finish this area up, then. In a worse-case scenario, let's assume that the schedule is way off and none of these nukes are going to come onstream as quickly as people say they are, in fact, we're well into 2004 and the problem is still there. Can you just give the Board a view of what that means for the forecast that you've made? What problem does that create? 1355 MR. ROGERS: Well, I guess even if the nuclear units did not come back on, without repeating myself, we've added 1,068, 10(6)m3 into the forecast for the new high-efficiency merchant gas- fired plant, 1,220 megawatts. Those are on, they are contractually committed; we know that. 1356 And we also know that even if the situation arose for the large merchant gas/oil-fired merchant plant to run, I checked my numbers since yesterday and got the -- I've clarified them. There is a differential of a dollar U.S., or depending on what rate you would use to convert that, $1.35 to $1.40 Canadian per gJ price advantage to oil on the forward curves, going forward, and we substantiate that commercially. So even if the units were available to run, they would be displaced by at least two-thirds of higher efficiency gas-fired cogen; and then, if they were available, they would be pushed to the very high peaks, and at that point the gas price that we would be trying to compete with would be $1.35 Canadian per gJ on the market. 1357 So I see a number of barriers. Certainly I'm pushing my salespeople to try to work through that, but there's not much light at the end of the tunnel. 1358 MR. MORAN: Thank you, Mr. Rogers. 1359 Now turning to you again, Mr. Gardiner, I wonder if you could just turn up Exhibit J.7.1, please. 1360 MR. GARDINER: Yes, I have it. 1361 MR. MORAN: This is the interrogatory that shows the breakdown of the revenue requirement that's been discussed at various points during your appearance here. 1362 Just so I understand how all of this works, the weather people do their thing and they produce a forecast of heating degree days which, as I understand it, they pass on to you so that you can do your volume forecast; is that correct? 1363 MR. GARDINER: Actually, I produce the weather-normal. 1364 MR. MORAN: Okay. So you do both of them? 1365 MR. GARDINER: Yes. 1366 MR. MORAN: So you develop the forecast for heating degree days and then you develop the volume forecast? 1367 MR. GARDINER: That is correct. 1368 MR. MORAN: And then in terms of figuring out what the revenue requirement is and trying to figure out if there's a deficiency or a sufficiency for the purposes of rate-making, is that passed on, then, to someone else by you? 1369 MR. GARDINER: That goes to someone else. 1370 MR. MORAN: All right. So that when we see the numbers that's set out in J.7.1, that's why we see Pat Elliott's name at the bottom, because she's taken your HDD forecast, put through the volume forecast, and said, Based on that, here's our revenue issue. 1371 MR. GARDINER: That's correct. 1372 MR. MORAN: And then specifically if we look at item 2, then, as I understand, when we see $27 million related to item 2, the component that comes out of the change or relates to the change in methodology, that's the 20.4 that we've been talking about. 1373 MR. GARDINER: Correct. 1374 MR. MORAN: All right. And does the stuff that you do with respect to predicting forecasting heating degree days and then the volume relate to item number 3 as well? Is that an input for item number 3, the 22 million? 1375 MR. GARDINER: The 22 million is based on the change in NAC on a common normal, the 2-04 trend normal between '99 -- as I mentioned yesterday, comparing the two years, and then multiplying it by the average number of customers in 2004, that gives you a volume; and then working with people in Pat Elliott's group, we use the average unit delivery commodity rate for each of the service classes. Those are -- I mentioned 7.3 cents the other -- earlier. That's the combined. But for each service class there's a -- for residential, it's around 8 cents; for the industrial, it's 4-something. And that gets multiplied through and you get this $22 million. 1376 MR. MORAN: All right. So if I understand it correctly, some component of the 22 million has to do with the change in methodology from 30-year rolling average to 20-year trend; is that right? 1377 MR. GARDINER: No, not in point 3. That is the change in normalized consumption. As we went from '93 to 2000. 1378 MR. MORAN: Normalized on what basis? 1379 MR. GARDINER: According to the 20-year trend. 1380 MR. MORAN: So if the 30-year had been used, this number might be different? In fact, it would be, wouldn't it? 1381 MR. GARDINER: Yeah, it would be different because -- and I don't have that number. It would not be the same number. 1382 MR. MORAN: Okay. That's fine. 1383 Just while you're on this point, you've done your forecast, and let's assume the actuals come right square with your forecast, you've got it exactly right. Am I correct, then, in understanding that the difference between using the 30-year methodology and the 20-year methodology is the $20.4 million? 1384 MR. GARDINER: That is correct. 1385 MR. MORAN: All right. So if you're exactly right and the actuals match the forecast, then $20.4 million more would have been collected in rates using the 20-year method instead of the 30-year method. 1386 MR. GARDINER: Yes. If the weather comes in at the 30-year average level as determined by the 1973 to 2-02 30-year period, if it comes in at that level, then there's an additional $20.4 million. 1387 MR. MORAN: Okay. And the reason for that is that 20.4 million has been built into the rates that are going to be charged on those volumes. 1388 MR. GARDINER: One point, though, is that the 27 is a comparison from the '99 rate case 30-year average -- it was the 30-years up to 1998. 1389 MR. MORAN: All right. I'll come to that point in a few minutes. Now let's look at the situation where your forecast is too high. 1390 MR. GARDINER: Okay. 1391 MR. MORAN: You're looking at -- well, let's start first with the too low. Let's say you're forecasting too low, we're talking about a cold winter in that situation and therefore an overearnings situation; right? 1392 MR. GARDINER: Right. 1393 MR. MORAN: All right. Given that the $20.4 million was built into rates on the assumption that this is the forecast and that's what we're expecting, to the extent that there's incremental volumes, those same rates, of course, will apply to those incremental volumes; right? 1394 MR. GARDINER: Correct. 1395 MR. MORAN: So to the extent that the $20.4 million has been built into rates and then applied to these incremental volumes as well, the benefit flowing from the cold weather, the overearnings benefit, is enhanced; right? 1396 MR. GARDINER: Yes, but it works the other way on the other side. And with symmetry, that's the risk. 1397 MR. MORAN: Yeah, fair enough. And if we look at the situation that happens if you overforecasted, as it turns out, we're looking at a warm winter, in other words, and in other words, underearnings; right? Again, because the 20.4 million has been built into rates, those are still going to be the rates that are applied to these lower-than-expected volumes; right? So to the extent that that is the case, the disadvantage of warm weather has been mitigated as a result. 1398 MR. PENNY: Well, if people aren't -- if people aren't consuming -- sorry, I just want to be clear about this. If customers aren't consume -- it's an assumption in your question. If customers aren't consuming the volumes, then they're not paying for them, and therefore the $20.4 million is not being recovered. And I think your question assumed that it would be. 1399 MR. MORAN: Do you need me to restate the question panel? 1400 MR. GARDINER: No. The other thing I'd like to add. I think you're getting into the realm of really how the 20.4 gets translated into rates and what the rate would look like under a 30-year, under a 20-year declining trend, and you're now at the margin. And I'm not the person that can sort of get into that margin detail on the rate. But I think that's where you're going and I really can't address that. 1401 MR. MORAN: And I don't want to lead you into an area that -- 1402 MR. GARDINER: No, I'm being quite frank with you. 1403 MR. MORAN: That's fair. Just to make sure that the question is clear, and then if you still feel it's better for someone else to answer, that's fine. 1404 To the extent that the 20.4 was built into rates on the basis that this was the forecast, this is what we expect, but in the situation where volumes less than forecast is the situation, to the extent that that was built into the rates, those are the rates, obviously, that would be applied to those lower volumes; and to the extent that that's the case, the fact that 20.4 was built into the rates will act as a mitigating influence. It won't be numerically 20.4, but it will reduce the disadvantage that's created by the warm weather; is that fair? 1405 MR. PENNY: Reduced as opposed to what, Mr. Moran? I guess that's what's unclear about that. 1406 MR. MORAN: Compared to the 30-year method that may be used instead. 1407 MR. GARDINER: I think I understand, but I think what you're getting at is this incremental difference in the rate depending on -- if you build the rates on the 30-year average, the rates would be like this. If you did the rates on the 20-year average, they'd be like that. And I don't know what they would look like, okay? 1408 MR. MORAN: Right. 1409 MR. GARDINER: And then you say, okay, if you had a 20-year, this is what the volumes look like, 30, and then if you price them, there's going to be a price differential difference and that's what you're getting at. I can't answer that question. 1410 MR. MORAN: That's all right. That's fine. And that would be the next panel, then, that I would raise that with? 1411 MR. GARDINER: It would be the rates panel. 1412 MR. MORAN: The rates panel, okay. 1413 Moving on to my next area, then, and, Mr. Chair, just so you can keep track, I only have three more areas, one of them very brief. 1414 I think Mr. Thompson was able to get on the record the other day that, on the basis of the 20-year trend method, the heating-degree-day forecast for the north was 4,953; for the south, it was 3,677; for a combined number of 3,996; right? 1415 MR. GARDINER: Yeah. 1416 MR. MORAN: And then later on it was pulled onto the record that if you used the 30-year rolling average method to forecast for 2004, the number would be 5,224 for the north; 3,919 for the south. I don't believe we have the combined number. Do you have that? 1417 MR. GARDINER: Yes, I do. Bear with me. I have that. 4,245. 1418 MR. MORAN: 4,245, thank you. This was a test of Mr. Wightman's math. He got the right answer. 1419 So on that basis, then, we see the difference, obviously, between the two methods from a forecast 2004 basis. 1420 If the Board were looking at the fact that a complete change from 30 to 20 creates this kind of a revenue deficiency, and try to look at what a transition period might look like, this is what I want to explore with you, and you take the two methods together and then weight them to produce an HDD forecast, would you be in a position to tell us what the forecast would look like and then the resulting revenue requirement change if you were to weight the 30-year method, 75 percent, against the 20-year method, 25 percent to produce a heating degree day forecast which you could then run through your volume forecast? 1421 MR. GARDINER: That could be a very large undertaking. 1422 MR. MORAN: And why would that be? 1423 MR. GARDINER: Well, the demand forecast part of it isn't difficult, but as we were discussing with Madame Singh, Mrs. Singh, this flows over into the financials and there's gas costs and all that, so it's quite a large task to do. In the interrogatories, on J.17.16, we had those NAC comparisons. To get the volume difference, you know your split, you know the NAC difference, you have the average number of customers, and the volume can be calculated that way. Now, getting into the revenue side of it, that's a little work. 1424 MR. MORAN: As I understand it, you take the HDD forecast, you run it through your process, your equations, to come up with the volume forecast. That part of it, I understand, would be quite straightforward. It's the revenue part, the subsequent part, that's a problem; is that what you're suggesting? 1425 MR. GARDINER: That's part of it that's -- yeah, that's what I'm saying, yes. 1426 MR. MORAN: All right. Well, why don't we take it, then, in two pieces and see where it takes it. Would you be able to produce the volume forecast using the weighted method, as I've suggested, to produce the HDD input for a 75/25 weighting in favour of the 30-year, and then a 50/50, and then a 25/75 in favour of the 20-year. 1427 MR. PENNY: Just so I understand, you're saying for one year, 75/25 weighting in favour of the 30-year average; then for a second year, a 50/50 weighting between the two methodologies; then for a third year, a 75/25 weighted in favour of the 20-year average? 1428 MR. MORAN: No, for the same year, for the 2004. 1429 MR. PENNY: All for one year? 1430 MR. MORAN: Yeah, a sensitivity analysis for 2004. 1431 MR. GARDINER: How precise do you want the number to be? 1432 MR. MORAN: As precise as it needs to be to be useful, Mr. Gardiner, to understand the differences. 1433 MR. GARDINER: No, because the math -- I mean, earlier today, in response to Mr. Aiken, on a per-heating-degree-day basis -- it could be done that way instead of running it through all the models, heating degree days... 1434 MR. MORAN: I guess I'm not asking you to do anything with the model that would apply to the HDD forecast. I'm just saying, take what was forecast for 2004 using the 30-year model and the HDD forecast using the 20-year model and then weight them as the input for the volume forecast. You'll be able to do that? 1435 MR. GARDINER: I can do that. 1436 MR. MORAN: Okay, thank you. 1437 Mr. Chair, that will become Undertaking M.5.5. 1438 MR. SOMMERVILLE: Thank you. 1439 UNDERTAKING NO. N.5.5: TO TAKE WHAT WAS FORECAST FOR 2004 USING THE 30-YEAR MODEL AND THE HDD FORECAST USING THE 20-YEAR MODEL AND THEN WEIGHT THEM AS THE INPUT FOR THE VOLUME FORECAST 1440 MR. MORAN: All right. The second bit, then, is what does that do to revenue requirement? And you're not sure how much work is involved in that. I'm wondering if the company would undertake to consider what's involved in producing the related changes in the revenue requirement, with those three scenarios, and report back. 1441 MR. PENNY: Yeah, I think it was Mr. Shepherd, probably, not Ms. Singh raised this, because I think this is along the same lines. An adjustment to the heating degree days for Mr. Gardiner's purposes is a more straightforward exercise. The problem comes when you take in all the revenue impacts of that: Gas supply planning, the transmission and the whole thing. But we'll do the same thing with Mr. Moran's question as we said we would do with respect to Mr. Shepherd's. 1442 MR. SOMMERVILLE: Thank you, Mr. Penny. 1443 MR. MORAN: Thank you very much. 1444 Okay. So then that takes me to my next area. Just to confirm, if I want to get into the nuts and bolts of how the numbers in J.7.1 are produced, that's not you, that's Ms. Elliott? 1445 MR. GARDINER: Right. 1446 MR. MORAN: That area was very fast, which takes me, then, to -- all right. Mr. Wightman has just scratched off another question, so this is all very positive, Mr. Chair. 1447 Mr. Gardiner, would you agree that for the heat-sensitive loads, the realized weather conditions are the major determining factor to be considered? 1448 MR. GARDINER: The realized? 1449 MR. MORAN: The actual weather. 1450 MR. GARDINER: Yes, the actual weather is the prime demand driver. 1451 MR. MORAN: This is a question of clarification. You were talking about the difference, the relationship between price and effect, the changes in price and the effect it would have on your NAC calculations. I think in your prefiled evidence, at C.1, tab 1, appendix A, you were talking about a 10 percent change and what impact that would have. Do you want to take a moment just to check that? Page 5 of Exhibit C.1, tab 1, appendix A. 1452 MR. SOMMERVILLE: 5 of 5? 1453 MR. MORAN: 5 of 5, that's right. At line 13 you indicated that a 10 percent change in price will change both residential and commercial consumption by approximately 1 percent? 1454 MR. GARDINER: Yes. That's approximately. As I mentioned earlier, in the residential market, the price elasticity is actually .7. I don't recall what the commercial one is for 30M hot but it's around 1 percent. 1455 MR. MORAN: And is that a short-run or long-run elasticity? 1456 MR. GARDINER: It's basically a short run, the next two years. It's not... 1457 MR. MORAN: All right, thank you. My last area, then, has to do with the issue of how you deal with heat-sensitive load. As a general proposition, as I understand it, a heat-sensitive load is going to be normalized for the purposes of your forecast? 1458 MR. GARDINER: The general service customers, with the exception of rate 16, and as I mentioned there are a small number of rate 10 customers that are not in my core market, so the ones that are in the core market, they're all weather normalized. 1459 MR. MORAN: All right. And the reason you do that is you get a better forecast if you take into account the sensitivity? 1460 MR. GARDINER: Right. 1461 MR. MORAN: When we look at Kitchener's load, we know that as a city they have a heat-sensitive load because that's the nature of that particular load; right? They have a strong -- 1462 MR. ROGERS: I'll take that one because Kitchener is my large account. I guess what I said yesterday was that we really don't know what the makeup inside of Kitchener is of the various kinds of customers, because they're a wholesale customer which take delivery at one meter and so be it. We just don't have the breakdown of what portion of their loads are heat-sensitive. I would grant you that they are residentials, which is a large number. Could be. 1463 MR. MORAN: Right. You don't know the breakdown, but what you do know is that within the Kitchener customer, there's this portion that's heat-sensitive; right? 1464 MR. ROGERS: I would grant you the fact that the residentials are heat-sensitive within the City of Kitchener. 1465 MR. MORAN: Has Kitchener asked you to normalize the heat-sensitive component of their load? 1466 MR. ROGERS: There's been an undertaking to illustrate that yesterday, and I can't remember the exact -- I think 80 percent was the number that Mr. Quinn or Mr. Ryder had asked us to do, and we're currently completing that undertaking as we speak. 1467 MR. MORAN: All right. If you get the breakdown and confirm it, would it be a problem to normalize that component of their load? 1468 MR. ROGERS: I'd have to consider that. I really ... 1469 MR. MORAN: It wouldn't be a problem, would it? 1470 MR. ROGERS: Mechanically. The issue is, I guess, the rationale behind doing that when we deliver to one meter. But mechanically I'm -- just intuitively, it should be able to be done. 1471 MR. MORAN: All right. So if they can give you the information you need, you'll be able to accommodate them with respect to the issue -- 1472 MR. ROGERS: Well, we don't normalize any wholesale customers. That was my point yesterday. We do not -- I believe we do not normalize exfranchise customers, Enbridge, Kingston. So the question, I guess, for the company would be beyond just the City of Kitchener, why would we step in and just do it there. So I don't know the answer to that. It's broader than my area of responsibility and accountability, so I think you've sort of moved outside to another panel, corporate panel. 1473 MR. MORAN: That's fine. Thank you, Mr. Rogers. 1474 Those are all my questions. I came in 10 minutes under my estimate, Mr. Chair. 1475 MR. SOMMERVILLE: The panel has no questions, Mr. Penny. 1476 MR. PENNY: Yes, thank you, Mr. Chairman, just a few questions in re-examination. 1477 RE-EXAMINATION BY MR. PENNY: 1478 MR. PENNY: In the context, Mr. Rogers, of Mr. Rattray's questions, who was representing OPG, there was a discussion from some of his exhibits, there was one in particular, I think, that dealt with the M7 forecasts, and you had talked in your answer about M7 and T1 rate-switching. Do you recall that? 1479 MR. ROGERS: Yes, I do. 1480 MR. PENNY: Can you first of all explain what you mean by rate-switching and give us some description of what, historically, that meant over the past two or three years as they appeared in that schedule? 1481 MR. ROGERS: Yes, I can. If we look at -- without referring specifically to the schedule -- if we need to go deeper, we certainly can. The wish was that unless we combine the M7, which is a firm bundled-T -- 1482 MR. PENNY: Sorry, Mr. Rogers. Just explain what M7 and T1 rate-switching is, and then we can move to having to look at them together. But explain what it is first. 1483 MR. ROGERS: Okay. What it means is, for instance, if volumes were forecasted to be burned under M7 but, in fact, the actuals were -- the rate-switching did not occur and the actual through-puts for that customer were experienced and reported under T1, then you have a differential. And rate-switching would have been forecasted to happen by a number of -- for a number of M7s to migrate to T1 service during the forecast period. And in the past we've had those discussions with a number of large customers. 1484 As I said previously, they have been reticent to move to the semi-unbundled T1 service, and in doing so, that's what created the variances in prior years. 1485 MR. PENNY: Let me break that down. You just mentioned that you have customers who say they are going to move from one rate class to the other; is that what happens? 1486 MR. ROGERS: That's correct. 1487 MR. PENNY: And if a customer does -- if a customer says that -- an M7 customer says, I'm going to move to T1, what are the implications of that for your forecast? 1488 MR. ROGERS: What it means is we would forecast that particular customer to burn under T1. And so if he did not make the move, then his -- we would be incorrectly -- let me just back up. Under the M7, if we predicted him to migrate to T1 and he burned under M7, then the actuals in M7 would be over. 1489 MR. PENNY: And the actuals in T1 would be? 1490 MR. ROGERS: Would be under. So without combining the two of them, you wouldn't have a picture where the volumes were for that particular customer. By putting them together, you remove that rate-switching and neutralize it and can see the true consumption differential for that customer, whether it be under a M7 or a T1. 1491 MR. PENNY: And to what extent does that issue, then, of customers who say that they are going to go to T1 and then don't, or perhaps vice versa, to what extent does that particular phenomena explain the forecast variability that was showing up in your M7 forecast? 1492 MR. ROGERS: Well, on the volume -- and we're completing an undertaking now; we're not finished yet, but if we do, I spoke to this yesterday, that virtually takes the number down to an average over the four years that were in that M.4.4 exhibit to an average of minus 4 percent, ranging from 0 to minus 1, minus 16, and 2. So it just completely removes the huge variability that you would only see if you observed just the one class, which would be the firm M7. 1493 MR. PENNY: Okay, thank you. Now, you also had a discussion with Mr. Rattray about an alternate oil/fuel generation plant. 1494 MR. ROGERS: Yes, that's correct. 1495 MR. PENNY: Just because it's hard to say, is there any reason why we can't talk about what plant that actually is? 1496 MR. ROGERS: Well, I've been using that -- we try not to use specific customer names during the proceeding. 1497 MR. PENNY: I'll -- 1498 MR. ROGERS: So everybody knows what it is. It's one of those -- 1499 MR. PENNY: If everyone knows what it is, then why don't we say what it is. 1500 MR. ROGERS: It's Lennox. 1501 MR. PENNY: That would make my questions a lot easier so I won't have to say the OPG alternate gas/oil -- 1502 MR. ROGERS: We've been talking about that for the last three days. 1503 MR. PENNY: What level of contractual commitment do you have from Lennox -- does Union currently have from Lennox with respect to gas consumption? 1504 MR. ROGERS: None. The original 15-year contract began -- that began burned -- they were fulfilled to their minimum -- maximum usage over a five-year period which ended August of 2003. So we have no contractual minimum commitment to burn, going forward. 1505 MR. PENNY: And you already told us, in answer to a question from Mr. Moran, about the oil/gas price differential. But in the context of OPG, you also said, in answer to some questions from Mr. Rattray, that you talked to your customers. What, if anything, have you heard from OPG/Lennox about its capability to burn oil at the Lennox plant? 1506 MR. ROGERS: We know that because of the price differential that I've spoken about before, the dollar U.S. per gJ, that OPG is, in fact, planning to burn oil this winter and are making the appropriate arrangements to do so and virtually, as far as we know, will be -- are fully loaded with oil in their tanks going into the winter season. 1507 MR. PENNY: All right. And then is there, Mr. Rogers, sufficient TCPL capacity to serve Lennox on a firm basis, because it's in the east end, in the winter of 2004 if it were going to be burning substantial gas to generate? 1508 MR. ROGERS: No, there is not. 1509 MR. PENNY: And what arrangements has Union offered to overcome that problem? 1510 MR. ROGERS: Union has been in discussion with OPG to offer a winter exchange service. 1511 MR. PENNY: And what has the OPG response been to that? 1512 MR. ROGERS: So far we have not received a positive response to that commercial offer. 1513 MR. PENNY: So they're not willing to enter into that exchange. 1514 MR. ROGERS: That's the result. 1515 MR. PENNY: And what does all that tell you about your forecast of the -- of gas consumption at the Lennox plant? 1516 MR. ROGERS: What it tells me is that during the winter period, we will have great difficulty to deliver physically, against a competitive oil price; and if the nuclear plants are back, that completes the triple play that would keep us from serving. 1517 In the summer, we see that while we could physically serve it, there is more likelihood, as you move through the period, that all of those factors will be in place: The new nukes, the new plants at Brighton Beach, and Imperial Oil, and the price, as we see it forecasted going forward, will still be $1.00 U.S. or approximately $1.35 Canadian per gJ out of the market. 1518 MR. PENNY: Thank you. Mr. Reghelini has just pointed out that I didn't quite tie the loop on the T1/M7 rate-switching. Do you have any expectations about whether the same level of T1 and -- whether this problem of customers saying they might but then don't switch, what circumstance is there today about the likelihood of that problem happening again in the future? 1519 MR. ROGERS: Yes. The changed circumstance for 2004 is that we've been talking to a number of large customers which is why that number is so big. And as I spoke about the issue of Union looking at changing some of our balancing provisions, three of our largest customers have signed contractual commitments, in effect, November 2003, and so they will be -- they are now contractually bound to move and migrate to T1, complete the switch, and that should take away a lot of the variability for 2004. 1520 MR. PENNY: All right, thank you. 1521 Mr. Poredos, you were asked by Mr. Janigan, you'll recall, about the change in the market hub transactional business, and the fact that you had found some -- I think it was some customers were using the market hub services incorrectly as a replacement for short-term storage and you sort of turned off the tap on that. Do you recall that exchange? 1522 MR. POREDOS: Yes, I do. 1523 MR. PENNY: And what would the impact be on customers being denied the market hub service as a replacement for storage service? What do they do, then, in that circumstance? 1524 MR. POREDOS: Well, what the customers did is use a cheap hub account to overload it and then they pulled it out at an interruptible rate, at a very low price. The issue is that those customers really shouldn't have been using that service, or that service for the purpose that they were trying to achieve. They should have gone and bought peak storage or off-peak storage, and that would have meant that they probably would have paid a higher rate. 1525 MR. PENNY: And is that what customers, in fact, did when you indicated that you're tightening up the requirements for market hub? 1526 MR. POREDOS: Yeah. The hub agreement has been changed to lower the limits of availability and that's been clearly communicated to customers, and they've accepted that, going forward. 1527 MR. PENNY: And so how have you reflected that in your forecast? 1528 MR. POREDOS: It's lower by about $1 million, I believe, because now you can't remove as much out of the those hubs. 1529 MR. PENNY: All right. And has that had any impact on your short-term storage service? 1530 MR. POREDOS: It would probably increase our short-term storage revenues with off-peak usage. 1531 MR. PENNY: All right. And then there were some questions about Houston and whether they were doing a good job down there. Have you seen any evidence that Houston salespeople are not giving Union as good service as Union used to get? 1532 MR. POREDOS: Absolutely not. 1533 MR. PENNY: And have you seen any evidence that Houston salespeople are not doing as good a job marketing excess capacity as Union people used to do? 1534 MR. POREDOS: I certainly have not, and I have not gotten any information from outside the company that would suggest that. 1535 MR. PENNY: And then my final question, I think, is for you, Mr. Gardiner, and it relates to some questions that came from Mr. Aiken - perhaps it's a similar issue arising out of Mr. Moran, I'm not sure - about the statistical issues associated with T stat and the reliability of those statistics. Do you recall that? 1536 MR. GARDINER: Yes, I do. 1537 MR. PENNY: And my question to you is simply that, looking at M.5.2, which is the backcast of your actuals to your forecast, that if the statistical issues associated with T stat or heteroskedascity or non-stationary, if those issues were of anything more than merely an academic interest, would you be getting this kind of match in 5.2? 1538 MR. GARDINER: I don't believe I would. 1539 MR. PENNY: All right, thank you. 1540 Those are all my questions, sir. 1541 MR. SOMMERVILLE: Mr. Penny. 1542 Is there anything further before we adjourn for today? 1543 MR. PENNY: I have nothing. 1544 MR. SOMMERVILLE: I'd like to thank the panel for its very diligent -- I know this was a pleasure from start to finish, and all good things must come to an end. I don't think any of you are scheduled to return, so have a safe trip back to Chatham, I guess, and have a nice weekend, and that is extended to all. 1545 MR. PENNY: As we to you. 1546 MR. SOMMERVILLE: Thank you. And we will adjourn until Tuesday morning at 9:30, and we will begin with the human resources panel at that time. 1547 Thank you very much. 1548 --- Whereupon the hearing adjourned at 4:50 p.m.