Energy Power Supply

Regional Power Supply Adequacy/Reliability Study
Status Report and Report on the Third Advisory Committee Meeting

February 18, 1999

The Advisory Committee held its third meeting on Thursday, February 18 in the Council's Offices in Portland. Committee members attending were: Ray Bliven, representing some aluminum companies; Steve Kerns, BPA; Ed Clarke, Ida-West; Tim Smith, BPA; Linc Wolverton, Industrial Customers of NW Utilities; Rich Lauckhart, NW Power Enterprises; Jack Williams, Oregon PUC; Steve Knudsen, PG&E Gas Transmission; Chris Elliot, Northwest Power Pool; Marty Howard, representing the Columbia River Inter-Tribal Fish Commission. Also in attendance were Carl VanHoff, Supply System; David Stewart-Smith, Oregon Office of Energy; Kurt Granat, PacifiCorp; Mike McCoy, Consultant.

Status of the modeling effort

The objective of the modeling effort is to be able to describe in probabilistic terms the magnitudes and frequencies of unserved load. Because we no longer can assume that if we have adequate energy over a period we will be able to cover peak loads with the hydro system, the analysis will go down to the hourly level as well as the more usual monthly analysis. Pete Swartz began with a discussion of the status of the modeling effort (GENESYS). He discussed his "punch list" for the key items remaining. The punch list includes:

• Implementing inclusion of non-treaty storage
• Including the most up-to-date firm export and import contracts data
• Incorporating stochastic variation of hydro capacity at a given energy
• Revised hydro Forced Outage Rates and updated capacity curves from Trapezoidal method.
• Improve thermal forced outage treatment
• Refine data assumptions (reference load, resources, transfer capabilities, Banks Lake)
• Integrate multi-dam hydro regulator
• Refine hydro pricing
• Reserves accounting
• Incorporating the ability to address changes in system configuration during the study

Pete then turned it over to John Fazio who described some of the work to model the hydro blocks in GENESYS. His presentation is attached. The key point of his presentation was that there is a relatively large block of hydro below contractual drafting rights that is not normally used because of fish and wildlife constraints. It may be possible to temporarily violate those constraints without adversely affecting flow requirements for fish later in the year. It will cost something to purchase energy to permit the reservoirs to refill after the constraints have been violated for reliability, but that cost could be considerably less than the cost of either curtailment or alternative means of maintaining reliability. Use of this block of hydro would be controversial. There was some discussion of whether use of this water should even be considered in this analysis. On the other hand, if infrequent use of a portion of that hydro block can improve reliability without adversely affecting fish flow requirements, it would seem to be a reasonable option. Staff subsequently decided to analyze the reliability situation without use of that hydro block and then, if necessary, evaluate the effect of using a portion of that block on the frequency and the magnitudes by which fish flow targets are not met.

Pete returned to describe the model structure and the underlying approach to modeling the hydro dispatch. The representation of the power system incorporated in the model is shown on Figure 1.

The power system is represented by 5 areas corresponding to the Pacific Northwest -- East (of the Cascades), the Pacific Northwest -- West, Northern California, Southern California and the Desert Southwest, and Canada. The transmission links between them are represented with transfer capabilities. The South to North Transfer capabilities vary as a function of net loads west of the Cascades (PNW West) consistent with Bonneville transmission nomograms. A dispatch of generation to meet load is performed for the PNW resources. Resources in the other areas are treated as a supply curve in an amount sufficient to fill the transmission links if necessary. Conservatively it has been assumed there is no import capability from Canada. The model will import from California when it is economic to do so. The focus of this study has been winter time reliability. In those months, California and the desert Southwest typically have significant excess generation capability. Consequently, it has not been felt that a more sophisticated representation was necessary. Comment was made that August can be a problem as well. A more complete representation can be incorporated if necessary.

Pete then gave a real time demonstration of the operation of the model. It was emphasized that THE RESULTS ARE ILLUSTRATIVE ONLY. The model simulated a series of Decembers, Januarys and Februarys, 2002 -- 2005 assuming very poor water (1937) and extreme weather conditions. The analysis assumed 1000 MW of new gas fired CCCTs in the Northwest and, as noted earlier, treatment of reserve requirements has not yet been incorporated. As would be expected, the model demonstrates periods with significant unserved loads under these conditions. Assumptions about the ability to rely on imports have the ability to significantly affect the results. The results of a several hundred game simulation in which water and weather conditions were allowed to vary stochastically were also examined. These illustrated the nature of problem -- the possibility of significant unserved load but with low probability of occurance.

There was discussion about the assumptions underlying the analysis. The fact that reserve requirements were not being treated adequately in the model was acknowledged. As noted earlier, treatment of reserves is on the "punch list". Similarly, it is understood that use of the lower blocks of hydro is by no means a foregone conclusion. It may prove that it is not possible to use those blocks.

The more difficult question is the treatment of imports and exports. The logic in the model says that if the Northwest is experiencing tight supplies and, hence, high prices AND there is power available outside of the region at a lower cost (including losses and transmission charges), that power will find its way into the region up to the limits of the transmission transfer capability. Similarly, if firm export commitments of NW utilities can be offset by purchases in the importing region (frequently referred to as counter-scheduling)at a cost less than the value of power in the Northwest, they will be.

Some members of the committee felt that those were not necessarily good assumptions. Concerns were raised at an earlier meeting regarding counter-scheduling. The concern was that unless firm transmission could be obtained to deliver the purchases made in the importing region, the purchaser might not allow its contract to be fulfilled through counter scheduling. They also cited the experience in last year's December cold snap in which there were few imports. California also experienced colder than normal weather and some gas fired generation was out of service because it did not have firm gas transmission rights. As a consequence, California prices were also high at a time when prices were high in the Northwest. This doesn't necessarily invalidate the logic in the model, but it does suggest that there can be circumstances in which power which would ordinarily be available for purchase from the Southwest might not be available. Staff will review what was going on during the December event to see what, if anything, it tells us with respect to the underlying logic of the modeling.

The point was made fairly emphatically that since the Bonneville Administrator is on record raising concerns about the large magnitude of regional deficits, we had better have a very good explanation if we come up with any different results. We should be able to duplicate the magnitude of the monthly and annual energy deficits from Bonneville's White Book if we make consistent assumptions about loads, new resource development, exports and imports.

Natural Gas

Steve Knudsen of PG&E Gas Transmission made some observations regarding gas price and supply. He noted that the gas prices we are using are reasonable for now. However, they may understate future prices. The primary reason is that our current price estimates assume that developers will not purchase firm pipeline capacity for all of their requirements, relying instead on cheaper, non-firm released capacity for some. Steve suggested that as the pipeline capacity market tightens, that may no longer be a good assumption and would push the transportation component of the gas price up. He also noted that during the December cold snap, several Northern California power plants were unable to run because they did not have firm pipeline capacity. He said he was unaware of any near term plans for expansion of gas pipeline capacity. However, "if you come, they will build it."

What Are We Doing this For?

The discussion then turned to the question raised by a short paper submitted by committee member Carol Opatrny (attached). Lon Peters submitted a memo in support of Carol's argument (attached). Unfortunately, neither Carol nor Lon were present to represent their views in person. At the almost certain risk of misrepresentation, Carol's thesis was described as follows:

Dick Watson attempted to clarify to objectives of the project. He indicated that the Council would continue with the modeling effort because these are improvements to the Council's analytic tools that we would want to make in any event. However, the richer understanding of the situation that the modeling can provide will, we think, be helpful in thinking about the policy issues.

With respect to the policy issues, the Council has never had any intent of making policy for reliability. Its intent was to provide information that would be useful to those who will be making decisions that can affect the adequacy and reliability of the region's power supply. To accomplish, we are not trying to predict where restructuring will come out. The intent is to try to carry out a fairly disciplined exploration of the options for assuring adequate supply and the pros and cons associated with each, probably as a function of position on the spectrum from regulated monopolies to retail open access. Carol has postulated one option -- maintaining the obligation to serve at the local level. There are others that probably should also be evaluated. If, for example, the region pursues the formation of an Independent (Transmission) System Operator (ISO), should the ISO have the responsibility for assuring adequate supply as is the case with the California ISO?

Linc Wolverton stated that he believed that restructuring was proceeding further than Carol seemed to assume and that it was not altogether clear that the local distribution utility was best equipped to address the adequacy issue. Others indicated that despite indications of potential deficit, no one was building anything.

Linc went on to suggest that the choices boiled down to insurance or the market. Insurance involves determining who has the responsibility and the ability to pay for keeping the "cats and dogs" (older, inefficient plants) around in order to assure adequate supply for infrequent adverse situations.

Regarding the question of the value of additional analysis, both the representatives of the developer community and the regulators indicated they felt that there was need for better information to help decision-making. The fact that analysis shows a large problem in the event of no new resource development, adverse hydro conditions and severe weather is attention-getting but not that useful.

The group decided to continue this discussion when Carol and others who shared her views could be present. In the meantime, work should proceed on the analytics to define whether or not and to what extent we really have a problem.

Resource Assumptions

The afternoon concluded with a presentation by Jeff King. Jeff presented data on plant availability; the availability of out-of-region resources that heretofore have been dedicated to Northwest loads (e.g., Colstrip); assumptions regarding the characteristics (heat rates, capital and fixed and variable O&M costs, permitting and construction periods); fuel prices and financing assumptions for new resources that could be developed in the Northwest. His presentation is attached. Comments were primarily directed to the issue raised earlier. To what extent can we expect the owners of plants with contractual obligations outside the Northwest to sell the output to the Northwest and satisfy their contractual obligation with purchases made in the Southwest when relative prices and availability of supply make it attractive to do so.

Next Meeting

The next meeting was set for Wednesday, March 31, 9AM to Noon at the Council's offices.

^ top