Rec. 11 - NW Energy Coalition

Response to the NW Power Planning Council's
Notice of Request for Recommendations: Mainstem Plan
(Document 2001-4)

June 14, 2001
Steven Weiss


The NW Energy Coalition ("NWEC") welcomes this opportunity to submit these recommendations. Our comments will focus on the relationship and proper integration of Program measures with the Council and BPA's obligation to assure an "adequate, efficient, economical and reliable power supply."

The current drought and power price "crisis" only highlight the well-known conflict between fish needs and power needs. It must be one of the Council's primary goals to reconcile and, if at all possible, mitigate this conflict. That, in our opinion, is a critical part of the charge set by Section 4(h)(11)(A) of the Regional Act which requires that Federal agencies provide "equitable treatment for?fish and wildlife with the other purposes for which such system and facilities are managed and operated." Unfortunately, the phrase "equitable treatment" is not specifically defined in the Act. But that lack of definition in no way lessens the responsibility of the Council to address the issue head on. NWEC believes that the questions asked by this "Request for Recommendations" are an important furtherance of that discussion.

The region's longer term goal should be to remove the health and welfare of the power system from the backs of Columbia and Snake River fish. Ultimately, fish can no longer be the "reserves" for the power system. The power system must pay its own way.

The Problem

NWEC sees the fundamental problem at the heart of the "equitable treatment" issue to be that fish needs are not currently a well-defined nor serious constraint on the operation of the power system. Because of this, fish operations and other expenses act as the shock absorber for both the physical reliability of the power system and for the financial health of Bonneville. Put another way, BPA's power business is not providing its own reserves.Instead it relies on fish for practically free emergency support.

Constraints on the power system, that is on the production and delivery of power, range from the very hard, to very soft and are physical, financial, social and environmental. Delivery cannot be made if generators can't generate enough, or distribution and transmission systems cannot physically deliver the power. These are very hard constraints. No matter how bad the crisis, we will not burn down the wires or destroy generators. Similarly, if there is no water behind a dam or fuel for a turbine, one cannot generate power. Physical hard constraints are self-enforcing and need no further discussion. Less hard, but more interesting, are financial constraints. If there are credit problems such that there is uncertainty of payment, or if prices are simply very high, one must think twice about purchasing or delivering power. Prices also dictate dispatch: which generator should run, how much water should be stored for later generation, and, for those with direct-access to the power markets, which end-use should be accomplished. That is, price determines whether or not power gets delivered and used. Environmental and social constraints include air pollution, recreation, navigation, flood control, and lost revenues from flow augmentation and spill for fish.

Some constraints are dealt with institutionally, while others are left to financial market forces. Historically, certain constraints have been treated as if they are very hard, while others are "monetized" to facilitate their trade based on financial value. Flood control and minimum stream flows for navigation do not receive a monetary value: instead these constraints are "hard-wired." That is, regardless of the value of or need to generate more electricity, the region will not push past those two constraints to threaten flood control or dry up the river. On the other hand, N0x and S0x have been monetized, so that the market, through the trading of pollution credits, determines whether it is worth producing additional pollution to create more electricity.

However, two important constraints have neither been monetized nor "hard-wired." These are the production of C02, and the value of spill and flow augmentation. Because neither of these constraints have any monetary value at present, nor are they subject to hard institutional constraint, they are controlled only through political pressure. If costs are equal, should we build a coal plant, a combustion turbine, or install conservation, given their different CO2 impacts? The only way we decide now is through the pressure of public opinion. And if coal is two mills cheaper? Then coal gets picked unless there is a political uproar.

We face a similar dilemma when it comes to hydro operations to help fish. If the region is unwilling to either make fish operations a hard constraint or give these operations an adequate monetary value, the only protection fish will have will be moral suasion and political pressure. NWEC believes the region and the Council can do better.

Another Problem

While we are pleased that the Council is seeking recommendations on how power and fish needs can be reconciled, actions being taken by Bonneville in its ratecase--to be finalized in a matter of days--may preclude many creative and sensible solutions. The Council will simply be too late, thus abrogating its responsibilities to the BPA Administrator. Therefore we have tried to make recommendations in that context.

For example, we have considered balancing fish and power needs through the use of a "Loss of Fish-Operations Probability," or "LOFP" to parallel the Loss of Load Probability ("LOLP") used in the Power Emergency Criteria. The idea would be that a projection of LOFP above 5% in the next 12 months would trigger a "Fish Emergency" that would automatically raise power rates, similar to how a LOLP triggers hydro deviations. However, we have not fleshed out such an idea because we believe that BPA's rates have been locked in and cannot be modified in this process.

Possible Solutions

BPA often asserts that it acts like a business, but it rarely does so. How would a business approach this problem? How do real businesses deal with unforeseeable, but very large risks? For example, how might a business plan for the possible loss due to unforeseen outage of a large piece of essential manufacturing equipment caused by breakage or even a natural disaster?

Basically, businesses mix and match solutions from several major strategies. I choose four most applicable to Bonneville.

A. Shift risk to customers. This entails planning for quick price changes without losing customers. Strategies can include locking in customers with long-term contracts, or creating niches in which there is little competition or product substitution available, so that the firm has the freedom to change prices quickly and substantially. Bonneville is depending almost exclusively on this strategy with its various CRACs.

B. Carry large financial reserves and/or untapped credit. BPA used to depend almost exclusively on this strategy. However, given the huge amount needed due to volatile markets, combined with weather risk, and the reluctance of Congress to provide more credit, customers have been unwilling to go this route.

C. Purchase insurance. Besides general policies covering fire, theft, and natural disasters, many businesses rely on "business interruption" insurance that can be written to cover very specific circumstances.

D. Maintain duplicate equipment. Physical "reserves" normally stand idle but can be quickly put into service. They can be also used to produce "non-firm" output when prices are high enough to justify the added variable costs of hiring temporary workers, obtaining short-term deliveries of feedstock, etc. These last two strategies have not been applied to fish issues as yet.

Each of these strategies has its advantages and disadvantages, but Bonneville has not explored them all fully, especially the last two. BPA's current strategy is to depend almost entirely on strategy "A". Unfortunately, this is a poor choice when it comes to fish and hydro operations. It requires the ability to very quickly adjust customers' rates radically. Because of that difficulty, BPA's rate staff and most of the customers have opposed proposed changes in the SN-CRAC which would allow it to really react quickly and robustly enough to purchase adequate spill for fish needs. We really can't blame them. Since NWEC represents consumers--especially low-income families--as well as salmon, we too hope a better way can be found to handle this issue in a more controlled fashion. However, since the region is relying heavily on strategy "A", we will spend some time discussing the changes which would have to be made to it to provide equitable treatment for both fish and power.

1. Strategy "A1" -- Declaring a Power Emergency vs. Triggering the SN-CRAC

On March 30 of this year, the Federal Agencies finalized three "Criteria" for declaring a Power Emergency for purposes of deviating from the new BiOp's spill and flow provisions. The first two criteria are dependent upon definitions of insufficient generation which ultimately depend on financial considerations. In almost all cases, water can be stored, load can be bought down, or power can be purchased to avoid outages--if enough money is brought to bear. The third criteria also is a financial criterion relating to BPA's reserve levels. Thus, the ability to trigger a power emergency--and thus deviate from fish operations--is ultimately almost solely a financial question. The only issue remaining is, how much money are the fish worth? (footnote 1)

This could be a necessary and honest debate, at least in the short term. (We hope that ultimately the region will no longer have to sacrifice fish in order to keep the lights on, and we believe that should be the Council's long-term goal.) Unfortunately, the way the criteria work--or at least the way the Federal Agencies have interpreted them--tilts the debate strongly against the interests of fish. In addition, the way the criteria work guarantees that the market won't get us out from under having this debate every time we have low rainfall or prices go up.

The problem with the Criteria is that they trigger much easier than the proposed financial trigger, the Safety-net CRAC, which would be used by Bonneville to reduce or eliminate the need for a hydro emergency at all. The Criteria depend upon a 12-month rolling forecast, while the SN-CRAC is limited to looking only to the end of each fiscal year. In addition, the Criteria trigger upon a finding that there is more than a 20% probability that BPA can't pay its bills, while the SN-CRAC requires a 50% probability to trigger. Finally, the SN-CRAC takes at least 5 months to go into affect, while the Criteria can change hydro operations instantly.

Therefore, since a hydro emergency can always be triggered before the SN CRAC can be implemented, it means that fish can be, and almost always will be, the first resort when financial times are tight. Once again the fish will have to depend political pressure rather than objective measures to participate in this debate. In addition, the generation and load-control market will never adjust to reduce the need for declaring a hydro emergency. If investors know that a hydro emergency can always be declared to reduce the need to spend lots of money to procure power and load curtailments, then they will never invest enough to reduce the frequency of this happening. After all, declaring a hydro emergency provides free (except for the political outcry) reserves for the power system. It's hard to compete with free.

It should be noted that the current Criteria represent a radical departure from the previously much more narrow definition of a power emergency which we quote in footnote 2. That definition much more specifically defined emergencies as physical emergencies. The current Criteria, on the other hand, create a mockery of the word "emergency," and instead are a mechanism to provide free financial reserves to Bonneville ratepayers.

  • We recommend that BPA's SN-CRAC trigger and the Power Emergency Criteria trigger be made compatible with each other. One can surmise from the Criteria that the Federal Agencies evidently believe they need at least 12 months advance warning to react in time to the possibility of insufficient generation. Bonneville must evidently believe that given the cost of failure, it needs a trigger sensitive enough to act on only a 20% probability of financial reserves being exhausted. These triggers seem reasonable. Thus BPA should change its SN-CRAC to conform.
  • We understand that BPA's final ROD may overlap and precede the Council's schedule. We have raised this issue in the ratecase, but the Administrator may not change his staff's current proposal. This then becomes a serious problem, because it limits the ability of the Council to instigate changes in the relationship between power production and fish restoration. The power side of the equation will have been set in stone. If that becomes the case, we urge the Council to put into place safeguards to ensure that Power "emergencies" are not triggered early merely to avoid the later triggering of the SN-CRAC. This is certainly anticipated in the Federal Agencies' Criteria paper which adds this caveat: "These planning criteria . . . will also take into account expected benefits of tools which are reliably available to mitigate cashflow problems . . . ." The Council should make it clear that before relying on a power emergency to deal with low reserves, Bonneville must use all of its available cash flow mitigation tools, including: (1) accessing its $250 million short-term Treasury note; (2) contracting forward for swaps or delivered power to be paid for after the SN-CRAC can be implemented; and, (3) imposing a transmission surcharge. Leaning on fish cannot be allowed to become merely a substitute for the provision of adequate reserves, both financial and physical, on the part of Bonneville's PBL.

2. Strategy "A2" -- Monetize the value of spill and flow deviations.

A good way to involve the market in reducing the frequency of hydro emergencies is to "monetize" the value of depending on an emergency. When there is a less than market cost to declaring an emergency (beyond the political fallout from environmental and fishing industry interests), the market, including BPA's load-serving utilities, will not efficiently develop alternatives. It will not be profitable or less costly to invest in peaker units and load controls if it is cheaper to lean on fish. There must be a real market cost put on deviations from the BiOp's recommendations so that the power system pays for its own reserves.

  • Sell the energy provided by deviating from the BiOp's hydro operations at a price incrementally--perhaps 10%--higher than its market value. Doing so will force the market to develop and choose lower cost alternatives. The money collected must not simply go back to BPA ratepayers, or the price signal will disappear. Instead, we recommend that the money should go to organizations such as the Bonneville Environmental Trust, to be used to augment habitat restoration, low-income assistance and weatherization programs and renewable resource development.

As we said earlier, "A" is not our preferred strategy. We do not believe it is the best way to deal with this issue, because it relies way too much on essentially "real-time" price adjustments which are very hard on customers and the economy. Because they are so radical, the region necessarily resists and takes the easy way out by leaning on fish operations. However that is the path that the region is on now. If the Council wishes to continue relying on this strategy, these proposals are necessary to protect hydro operations from the impacts of the power system.

3. Strategy "D" -- Acquire physical "Fish Reserves"

Eliminating fish operations altogether provides Bonneville with about 3700 MW-mos. of extra power spread over around 3-4 months, or about 1000-1200 MWs per month. This is a very unusual year, occurring only once every 25 years or so. That is the limit of what can be accessed in a Power Emergency. Let's get the power system's reserve requirements off the back of the river by acquiring those reserves directly.

  • Targeted load control, distributed generation and a limited number of simple-cycle CTs can be installed to provide the emergency reserves equivalent to that which can be captured from leaning on fish operations. A key to this proposal is that these resources must be kept available for use during the conditions now used to trigger a Power Emergency. They truly have to be fish reserves. And once built, fish operations could only be deviated from under the most stringent, physical emergencies, never for price reasons. Capital amortization and interest plus gas storage and/or firm gas transportation would be spread out over 10-20 years and made a part of annual fish program costs.

A rough estimate of the costs is as follows. Fixed costs for a simple cycle CT are about $47/kw-yr. Spill takes perhaps 1200 MWs, so $56.million/year. Drought conditions like this year produce gas spot prices of about $5.00/mmu. At a 10,300 heat rate that works out to about 51 mills/kwh plus variable O&M of 1 mill/kwh. It is not necessary to buy firm gas for spring and summer use. On average, under the conditions outlined in the Criteria--drought and very high prices--it might run about 300-500 hours/year times 1200 MWs, or 360-600 million kwhrs, for a variable cost of $19-31 million/yr. Total cost would be around $75-87 million per year, or around one mill/kwh to BPA's customers.

4. Strategy "C" -- Acquire financial "Fish Reserves"

This is our preferred approach. Instead of Bonneville actually acquiring physical reserves, the agency could buy "insurance" to cover the risk of conditions which now are used to declare a Power Emergency.

  • BPA should competitively acquire reserves from the market to cover the conditions requiring a hydro emergency. This could be a 10-yr. contract covering the subscription period, or even longer. The contract would define the conditions under which the reserves could be used (footnote 3). Bidders could offer peak generation, load curtailment contracts or simply financial hedges or insurance policies to fill this product. The advantage over strategy "D" being that the market would develop the most efficient way to serve this need. The annual cost would be similar to that estimated for physical reserves, the only difference being that someone else would manage the risk. The cost would be included in fish program costs.


We have approached this as a business would. Our principle has been that the power system should pay for its own reserves and not lean on fish operations. The cost would be quite reasonable when measured against the value of guaranteed spill for migrating salmon. It is the Council's duty to make sure that the declaration of a Power Emergency does not simply become a cheap and convenient way of providing reserves for the power system. We welcome this discussion and look forward to developing these and other proposals in a constructive fashion.


Footnote 1: We hasten to add that there are true physical emergencies which cannot be solved with money, and we do not object to the declaration of an emergency in those narrowly defined circumstances. However even these types of emergencies can be reduced in the long run if strategies of type "C" and "D" were utilized.

Footnote 2: From the 2000 TMT Emergency Protocol: "It is appropriate to define emergencies as they apply to the operation of the FCRPS. As evident from previous actual events, emergencies are a unique situation having the potential for many types of impacts, generally requiring some type of action or response to minimize or eliminate impacts. An emergency may involve the need to operate the FCRPS outside of the requirements contained in the 1995 Biological Opinions or the associated Records of Decision (ROD) issued by the operating agencies. These events may increase fish mortality above levels in the 1995 Biological Opinion and RODs.

"However, it is important to distinguish emergencies from "planned risks." In operating a complex system such as the FCRPS, certain risks are assumed every day. Future conditions are uncertain. Operational decisions rely on predictions, forecasts and probabilities. If an extreme circumstance occurs, it is not necessarily an emergency even though it was sudden and urgent, and caused an immediate action to be taken.

"For this protocol, emergencies are categorized into three types. They are restricted to power-type emergencies only. Each type is described below and illustrated with several examples.

1. Generation Emergency - the potential for or actual insufficiency of electrical generation to satisfy electrical demand or load in a particular geographical area. The insufficiency can be of short duration (a capacity shortfall) or have the potential to persist for a period of time (an energy shortfall) and is usually spread over a defined geographical area as determined by the interconnectivity of the transmission and distribution system. For example, a generation emergency may be caused by an unanticipated loss of a generating resource - a project/unit forced outage; or by a restriction in the amount of water available for project discharge - reducing on-site generation; or by a loss of electrical transmission capability used to import electricity into a particular geographic area - a transmission line restriction or shutdown.

2. Transmission Emergency - the potential or actual loss or limitation in the ability to move electricity from the site of generation to the actual consumer or end-user. For example, a transmission line may fail, shutdown or otherwise be unavailable to transmit any electrical energy - a line outage; or a physical condition may exist that prevents or limits effective and reliable transmission - - insufficient reactive power (VARs) to overcome the inherent losses in long-distance transmission; or a temporary limitation on transmission line capability that restricts the export of electricity - which causes a generation surplus in one area, thus reducing overall generation levels but causes a shortage in another area as noted above in the description for a generation emergency.

3. Other Emergency - the existence or result of extenuating circumstances which fall outside the range of normal operations, was unanticipated, and may have resulted in catastrophic impact, physical damage or failure to part of the physical power system. For example, all natural disasters fall under this category of emergency - earthquakes, floods, and fires; or human caused failures - ship or barge strandings, facility failures (e.g., locks, gates, outlets, etc.), chemical spills into the river, train derailments impacting the river and terrorist acts; or overriding circumstances or needs that require operations to exceed normal limits such as a police investigation, a rescue operation, and a project operation specifically designed to prevent damage to or protect other parts of the FCRPS."

Footnote 3: The RFP would, for example, cite the Criteria under which the reserves could be called upon. We predict that once monetary consequences must be accounted for, those Criteria would have to be defined much more specifically. As written they are exceedingly vague and open to arbitrary interpretation.