The Long and Short of Grid Congestion


FTRs make hedging possible, but can PJM ensure full funding without playing favorites?

Fortnightly Magazine - March 2016

Let's talk about financial transmission rights. But it's going to get technical, so hold on tight.

FTRs were the subject last month, when the Federal Energy Regulatory Commission (FERC) held a technical conference to review two tariff amendments that PJM proposed late last year to ensure that FTRs can continue to provide a viable means, as they were designed to do, of hedging against the risk of costly grid congestion.

PJM's twin proposals, now pending at FERC, attack a problem seen in past years known as FTR underfunding. That's what happens when the sum total of congestion charges that PJM collects when it clears the day-ahead market proves too low to cover PJM's obligation to pay off all the market players who hold FTRs as a hedge against that congestion.

PJM's solution would impact both types of financial transmission rights. It would affect both "prevailing flow" FTRs, of the type that utilities typically buy, which represent a "long" hedge; and 2) also "counter-flow" FTRs, which we might best describe as a "short" sale of congestion, and which are purchased more often by financial traders that owe no load-serving obligation. (See FERC Dkts. EL16-6 & ER16-121, filed Oct. 19, 2015.)

Of course, FTR underfunding may well prove impossible to eliminate entirely. FERC itself acknowledges that markets are unpredictable - that no one can design a financial hedge that guarantees protection against all types of risk.

Financial traders believe PJM’s proposal discriminates since they are more likely to hold counter-flow FTRs. –Bruce Radford

Yet now we see something new. With PJM stakeholders unable to agree on formal tariff amendment to solve underfunding, PJM's grid operators and engineers lately have undertaken ad hoc efforts to minimize the problem. In simple terms, PJM operators recently have revised certain engineering assumptions, such as transmission availability, the likelihood of outages, and the likelihood of interchange transactions with neighboring regions, such as MISO or New York, and so forth. With these tweaked assumptions, which essentially adopted a more conservative view of existing grid capacity available to support FTR hedging positions, PJM changed the mathematical analysis that tests for "simultaneous feasibility." And simultaneous feasibility is what helps govern the official forecast of how many FTRs can be issued and sold without breaking the bank.

But these tweaks have come at a cost. PJM's changed engineering assumptions in fact have created winners and losers, making more difficult for load-serving utilities (LSEs) to hedge congestion risk. So in one sense, PJM's new tariff proposals filed in October seek to undo some of the damage caused by the recent ad hoc attempts at a fix.

Can PJM find a way to assure revenue adequacy for FTRs without playing favorites among market participants? The utilities generally favor PJM's new proposals (as do state regulators in Maryland and New Jersey), but financial traders see red, fearing they'll become second-class citizens in PJM's wholesale day-ahead energy market.

FTRs: A Primer 

Financial Transmission Rights mark the second of two key concepts - the other being locational marginal prices (LMPs) for generation - that underpin the workings of the competitive wholesale power markets that FERC has approved for various regional transmission organizations (RTOs), such as PJM, California, MISO, New York, and ISO New England.

Locational pricing provides a way for transmission service rates to be expressed in terms of generation cost. That allows market players to ignore pancaked transmission rates and instead to "buy through" the congestion. It's what gas pipeline companies would call the "basis differential" - another term for the locational price premium commanded by generation that can be delivered to individual nodes on the grid.

So when PJM and other regional grids devised their day-ahead energy markets they largely did away with the contract-based physical transmission rights that utilities held to support their duty to serve native load. In their place, PJM provided these LSEs with FTRs. To be more accurate, what PJM did was to set up auctions to sell FTRs to competitive bidders, and then took the revenues from those auctions - known as ARRs, or "Auction Revenue Rights" - and allocated them to utilities. These ARRs are meant to make utilities whole against congestion risk. Or, if utilities prefer, they may exchange their ARRs for FTRs. (For our purposes, we may think of ARRs and FTRs as roughly equivalent in concept.)

Thus, FTRs fill a gap. They are meant to ensure that LSEs receive a stream of payments (reflecting transmission congestion credits) to offset the benefits the utilities lost when they surrendered their physical transmission rights as a necessary precursor to setting up the day-ahead market.

And it's wonderful when it works. But in truth, it's a matter of both guesswork and faith.

The guesswork comes when we endeavor to predict today how the grid will perform tomorrow when markets clear. Today's prediction governs the issuance of FTRs. Tomorrow's future governs how many revenues are actually earned. We hope they will agree.

The faith part is founded on a mathematical proof confirmed by Harvard Professor William Hogan.

Hogan proved that if two conditions could be satisfied, then any given array of FTRs plotted over a particular set of transmission paths and then sold via auction to regional market participants would always be revenue-adequate - that revenues from auction receipts would be sufficient to pay off obligations owed to FTR holders. First, the grid operator must confirm beforehand that the FTRs, in aggregate, are simultaneously feasible (that power can be moved on the grid along the paths defined and implied by the aggregate set of FTRs). Second, the grid topology must be the same both for the original issuance of the FTRs, and for the configuration that provides the basis for clearing the market through the RTO's security-constrained, bid-based economic dispatch.

But RTOs live in the real world - not the virtual world of PhD economics. Stuff happens. Like transmission outages. Chunks of intermittent solar or wind power sometimes show up where they were never expected. And real-time balancing of supply and load can create what's known as "negative congestion" - congestion in real time that runs opposite to the day-ahead bidding.

In fact, FTR revenue adequacy in PJM began heading South during the 2009/10 plan year (June to June) and continued to fall short for the five-year period ending with the 2013/14 plan year, before climbing back into positive territory during the plan year that ended last summer and during the first four months of the current plan year.

On why FTR funding fell short during the five-year period June 2009 through June 2014, the consensus seemed to point to negative balancing congestion appearing in the real-time market. That could happen, for example, if a transmission outage occurred during the time internal between the day-ahead market clearing and real time balancing, such that a given locational price differential between source and sink that occurred in the day-ahead market was effectively erased in real time. (For more detail, see our prior column: "Congestion on Trial: PJM and the crisis over FTR underfunding," Pub. Utils. Fortnightly, May 2013, p. 16.)

What Went Wrong

PJM's problems with FTRs over the past several years stem from two basic causes.

First, the region has seen large amounts of what is called negative balancing congestion. That means that the direction and/or magnitude of congestion that has shown up in the real-time market (the balancing market) has diverged frequently from what prevailed in the day-ahead clearing.

Second, PJM years ago committed to a level of ARR allocations for utilities holding long-term transmission rights that far exceeds simultaneous feasibility, but which PJM has no choice but to honor. And to honor those infeasible ARRs, known as "Stage 1A" ARRs, PJM perforce must cut back on the other ARRs (Stage 1B and Stage 2) that it allocates to LSEs to cover shorter-term rights.

Negative Balancing Congestion. How does congestion become negative? The direction of congestion ordinarily is positive. That's when prices at the sink exceed those at the source. And the positive FTRs purchased to hedge against such congestion are "prevailing flow" FTRs. But when direction reverses in the real-time market - that is, when prices at the sink fall below those at the source, or, which would be much more common, if the day-ahead price differentials between sources and sinks fail to achieve the same magnitude in real time as in day-ahead - then you have negative balancing congestion.

Or, as one financial marketer has put it, "negative balancing congestion ... derives from either reductions in transfer capability in the real-time market compared to the day-ahead market or from interregional coordinated transmission facilities that are not included [as part of the grid topology] in the FTR and day-ahead markets." (See, Comments of Appian Way Energy Partners, LLC, p. 3, FERC Dkt. EL16-6, filed Nov. 9, 2015.)

To better appreciate the problem, consider that of the $1.41 billion in FTR underfunding that occurred in PJM markets during the four planning years stretching from June 2010 through May 2014, some 90 percent of that ($1.28 billion) was negative balancing congestion. (See, Comments of J.Aron & Co., pp. 17-18, citing PJM statistics, filed Nov. 9, 2015.)

And why the problem? As it happens, PJM market rules require that real-time congestion revenues (in this case negative) are added back into the pool of money that defines how much funds are available to pay off holders of positive FTRs. The more that balancing congestion turns negative, the more that FTRs become unfunded.

Others concur: "As long as real-time balancing congestion is included in the calculation of available funds for FTRs, severe weather and other unpredictable events ... could in turn cause significant FTR underfunding." (Protest of DC Energy, LLC, Inertial Power, LP, Saracen Energy East, LP, and Vitol, Inc. p. 9, filed Nov. 9, 2015.)

Market Power & Market Manipulation - PUR Books

Long-Term ARRs. For the second cause, relating to infeasible ARRs for PJM's Stage 1A allocation, perhaps we should put some of the blame on Congress.

A decade ago, in enacting the Energy Policy Act of 2005, lawmakers added a new section 217 to the Federal Power Act (encoded in sec. 1233 of the text of EPACT 2005) by which Congress meant to give utilities a way to protect contractual long-term physical transmission rights needed to serve native load, or at least to ensure compensation to utilities for those rights by forcing RTOs to provide long-term FTRs as a hedge. FERC later translated the new law into tariff requirements for RTOs. (See, Order No. 681, Dkt. RM06-8, July 20, 2006, 116 FERC ¶61,077.)

This requirement has always proved problematic, as in theory it requires planners when they issue these long-term FTRs to anticipate how the grid will be configured years into the future, when the FTRs are settled or liquidated, or else give up on any faith (per Prof. Hogan's proof) that they will be adequately funded.

PJM solved this problem with a heavy hand. It decided simply to guarantee a certain allocation to LSEs of long-term, ten-year (Stage 1A) ARRs, based on historical patterns of supply and load attributable to those utilities, regardless of whether those ARRs are simultaneously feasible in terms of grid topology.

But at the same time, the TOTAL amount of ARRs that are allocated (Stages 1A + 1B + 2) must still satisfy a mathematical test of simultaneous feasibility. As a result, PJM ends up allocating many infeasible Stage 1A ARRs.

PJM explains: "[I]n allocating Stage 1A ARRs regardless of feasibility, PJM over-allocates ARRs, which exacerbate[s] the revenue inadequacy problem. (See, PJM: Proposed Modifications to ARR and FTR Provisions, p. 7, FERC Dkt. EL16-6, filed Oct. 19, 2015.)

And every one of those geographically infeasible rights comes at a cost: it means that fewer shorter-term rights can be issued in Stages 1B and 2.

In fact, the numbers show that from 2010 to 2015, the rate of allocations of Stage 1B ARRs (in response to requests) was reduced by over 90 percent - from 27,850 MWs in the 2010/2011 planning period, to 2,390 MWs during the planning period for 2014/2015. (See, Comments of Dayton P&L Co. & FirstEnergy Serv. Co.,
p. 2 (citing PJM statistics), filed Nov. 9, 2015.)

As was noted by Dr. Andrew J. Stevens, Managing Director of DC Energy LLC, in an affidavit he filed to support the DC Energy's protest (cited above), those numbers would suggest "that LSEs were seeking hedges on particular paths at quantities that exceeded ten times the actual capacity of the underlying transmission system."

Moreover, as was pointed out by three Public Service Electric & Gas Co., and its affiliates PSEG Power LLC and PSEG Energy Resources & Trade LLC, this denial of requested ARRs might well favor those PJM utilities situated in the western half of its footprint, forcing consumers in the eastern portion of PJM to pony up a subsidy.

To bolster that claim, the PSEG companies in their FERC filing cited findings in the 2014 State of the Market Report for PJM as prepared by region's Independent Market Monitor, Monitoring Analytics LLC (Joseph E. Bowring, President). In that report, the IMM had concluded that the greatest number of infeasible Stage 1A ARRs and the most severely constrained paths were MISO flowgates or lines located in the western portion of PJM. (See, Comments of PSEG Cos., pp 3-4, filed Nov. 9, 2015.)

The Planned Fix 

Now PJM comes forth with two very technical proposals.

First, PJM says it again will tinker with engineering assumptions - this time imposing a hypothetical assumption in its Stage 1A ARR allocation metric to reflect an additional 1.5 percent yearly load growth, which, according to PJM, should suffice to trigger requirements for transmission upgrades more quickly. That, PJM claims, should lead to more upgrades qualifying through its RTEP process ("Regional Transmission Expansion Projects), thus making more ARRs feasible more quickly for Stages 1B and 2.

Second, it proposes to end the practice of allowing FTR holders to credit their shorts against their longs before being exposed to FTR underfunding. To translate, that would mean no more use of counter-flow FTRs to offset a positive position on prevailing flow FTRs.

Let's take the second idea first.

Understand that funding shortages (and surpluses, for that matter) apply only to positive, prevailing flow FTRs. In other words, when you sell congestion short (when you bid successfully on a counter-flow FTR) you receive a payment equal to the absolute value of the market-clearing price on the auction of positive FTRs, but then remain liable only to pay back to PJM the eventual market-clearing price on day-ahead congestion for the same path. You are gambling that congestion will turn out to be less than others expected. But you are not gambling on the eventual revenue adequacy of positive FTRs.

Because of this fact, PJM traditionally has allowed market players to credit counter-flow positions against positive positions to calculate a net positive position - which is liable for underfunding or entitled to any surplus. But now no more, according to PJM's recent tariff proposal. In defense of this change, PJM says it should improve overall revenue adequacy for FTRs, since short sellers (those holding counter-flow FTRs) no longer will be able to use their positions to escape a portion of their liability for funding shortages.

Financial traders argue that this proposal plays favorites - that counter-flow FTRs improve the way power markets function (as some would say that short-selling does on Wall Street). They believe that PJM's new proposal discriminates against them since they are more likely than load-serving utilities to hold counter-flow FTRs.

But the PSEG companies point to analysis by PJM's IMM that if netting within portfolios as between positive and counter-flow FTRs would have been eliminated previously, then the payout ratio for FTRs over the 2012-13 planning period would have climbed to 84.6 percent versus the reported 67.7 percent. Ad they add that there's a reason why financial traders hold the bulk of counter-flow FTRs (80 percent, versus 20 percent purchased by entities with physical obligations to serve load, according to the IMM's 2014 State of Market Report). According to the PSEG companies, LSEs "may lack the ability to compete with sophisticated financial market participants that specialize in activities related to FTR/ARR ownership." (See, Comments of PSEG Cos., pp. 6-7.)

But the first proposal, to tack on an added 1.5 percent to the assumed yearly growth rate, sounds a bit more mysterious. To this reporter, it recalls Albert Einstein's infamous "Cosmological Constant" - a jiggering of numbers thrown in to the calculation at the last minute, almost as an afterthought, to make the numbers come out right, which the famed physicist was later to call his "biggest mistake."

In a claim that perhaps was better left unsaid, PJM wrote in its tariff proposal filing that its artificial 1.5 percent increase in assumed zonal base load "is not too aggressive to cause potential overbuilding of the system." Talk about faint praise.

A Shifting of Dollars

The fix that PJM declined to propose, but which wins mention most often from commentators on various sides, would involve passing the cost of negative balancing congestion not to FTR holders, but to the LSEs themselves - and thereby to consumers.

When FERC denied rehearing and eventually decided to do nothing about FTR underfunding in the big complaint case that was the subject of our prior column published in 2013, it endorsed PJM's method of setting off negative balancing congestion against FTR revenues, and flatly rejected the idea of pushing that cost on load-serving utilities:

"We continue to find that allocation of real-time balancing congestion to current FTRs has a reasonable basis, because FTR holders are in the best position to reflect the associated underfunding in the value of FTRs." (See, Order Denying Rehearing, June 8, 2015, FirstEnergy Solutions Corp. v. PJM, Dkt. EL13-47-001, 151 FERC ¶61,205.)

But to this reporter, that sounds like political expediency, a position taken to avoid controversy.

In fact, in its latest tariff proposal, PJM conceded that stakeholders "were discussing" changes to the allocation of balancing congestion - whether to spread the resulting uplift [charge] more broadly than to FTR holders." But, as PJM added, it chose not to "push for those reforms" because "it did not feel it was PJM's place to advocate a solution on which the stakeholders could not reach consensus."

After all, as PJM continued, the real solution, at base, would represent "a shifting of dollars from one group of market participants to another group." (See, PJM Tariff Proposal, pp. 10-11.)

And so both the Maryland Public Service Commission and the New Jersey Board of Public Utilities have come down hard against any suggestion of pushing the cost of negative balancing congestion on consumers, stating clearly that "end users are not the cause."

Rather, as the regulators have alleged, "this PJM action principally impacts LSEs and improves FTR market returns for financial market participants. ... End users are thus the principal injured party of this PJM action." And this position taken by the regulators from Maryland and New Jersey echoes the view of OPSI, the Organization of PJM States, as expressed in a letter sent via email to PJM in March 2015. (See, Corrected Motion for Leave to Answer of Md. PSC and NJ BPU,
p. 4, filed Nov. 27, 2015.)

So where does PJM really stand? The technical conference held at FERC on February 4 showed us a tantalizing glimpse.

Consider this exchange that occurred in the second morning panel, between David Patton, of Potomac Economics, who serves as the internal market monitor for New York MISO, and ERCOT, and as external market monitor for ISO New England, and Stu Bresler, of PJM:

PUF Books - New Regulatory Finance

David Patton: "Balancing congestion - I'd be interested in the rationale for allocating that to FTR holders, because it has nothing to do with how many FTRs are issued. There's no basis for it, and nobody does it other than New England, who adopted PJM's rules."

Stu Bresler: When the PJM market was designed, during the stakeholder process, the assumption was that balancing congestion could be either positive or negative, but would never really be all that large. And therefore we did not need to come up with a cost allocation mechanism for a separate congestion bucket, but rather, we would wrap all the congestion together and equalize it among FTR holders.

"Frankly, PJM's thinking has changed over time on that, and I was going to get to that later."

Lead image © Can Stock Photo Inc. / khunaspix