5 – PJM’s Black Box

If you read the PJM engineers’ testimony in Appendix H of the PATH application, you will see lots of references to PJM’s predictions of various “NERC violations” and reliability problems.

NERC is an agency that has set a single set of standards that all power operators must now meet during the operation of their transmission facilities.  A NERC violation occurs when transmission equipment begins to operate outside NERC limits for the temperature of the wire or electricity flows through certain circuits and transformers.

In Appendix H, we see PJM engineers showing us all the NERC violations that will appear in 2014 if PATH isn’t built.  All of these dire predictions come out of the “black box” of PJM’s “reliability modeling.”  PJM’s managers love to show us all the terrible things that come out of their black box, but they don’t want to show us what goes into it.

A computer-generated model, like PJM’s black box, makes predictions based on certain equations that include a number of variables based on past experience.  The equations themselves, and some of the variables, are based on a large number of assumptions about how past experience will produce future results.

PJM’s managers describe in great detail everything that comes out of their black box as if their results were facts that are certain to happen five years from now.  The problem is that the PJM managers don’t want to tell us much about the assumptions that go into their black box.  Even slight changes in certain assumptions could generate no NERC violations at all.

Here is what Robert Driscoll, Mirant Mid-Atlantic CEO, had to say in 2007 about PJM’s assumptions used to justify PATH (what Mr. Driscoll called the A-K Project):

Lastly, the transmission line loading models used by PJM staff for determining the need for the A-K Project by 2012 are highly sensitive to projected generation retirements and additions. PJM staff, however, has not released the assumptions used in the models so there is no way for market participants to test the validity of the results. In addition, PJM staff failed to follow their own guidelines by not evaluating in the transmission line loading models generation additions with proposed in-service dates that are prior to 2012. Thus, the A-K Project may not be needed by 2012.

In fact, Mr. Driscoll was more accurate than PJM’s managers in 2006.  Experience in 2007 and 2008 showed clearly that PJM’s original PATH start date of June 2012 was just wrong.  The real world proved that the assumptions that went into PJM’s black box were not very accurate.  Despite PJM’s claims that it operates in a “transparent” manner, PJM will not release key assumptions that go into its black box so that its accuracy, as Mr. Driscoll points out, can be tested.

Here are some things we do know about PJM’s general assumptions from the PATH application:

  • PJM claims the duration of the recent drop in overall demand in PJM will be “short.”  PJM managers offer no evidence as to why they believe that new residential construction and industrial production will begin expanding within the next year.  Most economists in the US are not so optimistic.  We have no way of determining if PJM is right or wrong if they will not provide the reasoning behind their rosy economic predictions.
  • In its original planning for PATH, PJM has assumed that DSM and increasing efficiency would not result in any reduction in the growth of peak demand in the PJM region.  PJM attributes all recent declines in peak demand to “the economic downturn.”  This is not surprising, because PJM only began its own Energy Efficiency (EE) program in 2008, despite the fact that PJM continually brags about how it is the oldest Regional Transmission Organization in the US.  While claiming that no other power companies’ EE programs will have any effect on peak demand, PJM now claims that its own EE program, only one year old, will have an effect on peak demand in the future.  PJM’s assumptions about the real and potential impact of DSM and EE are inaccurate and confused, at best.  PJM’s last major electricity auction showed a dramatic increase in sales of DSM resources.
  • In the past, PJM has not included any new electrical generation capacity in its black box unless the new capacity is a new power plant that has reached a certain point in PJM’s own application process.  In fact, modifications and additions to existing plants, particularly new natural gas fired equipment, can add significant capacity to the PJM process even without adding new plants.  This expansion process has resulted in lots of new capacity on the east coast that PJM’s black box failed to predict.
  • PJM’s past assumptions about how quickly existing plants would stop operating have also proved to be inaccurate.  In his 2007 comments to PJM, Mirant’s Robert Driscoll noted that PJM’s assumptions included shutdown of a major Mirant plant, although Mirant had not released any plans to close the plant.  That plant continues to operate today.
  • PJM also claims that important parts of its black box assumptions are private property of either PJM or the private power generators or retailers, and cannot be made public.

In the TrAIL case, the WV PSC was distinctly uninterested in looking into PJM’s black box.  While the PSC refused to accept expert testimony from intervenors’ witnesses about how bad PJM’s assumptions were, particularly about the impact of DSM and new generation capacity, the PSC bought all of PJM’s faulty assumptions about just those impacts.

The PSC has not commissioned an independent evaluation of PJM’s predictions.  PJM hired its own company to do an “independent” evaluation of its peak load forecasting.  I find it unconvincing that you can hire a company and its evaluation will be “independent.”  We have seen how that worked out with the “independent” rating agencies in the financial meltdown.

We need to be sure the PSC does its job in the PATH case.  That starts with requiring PJM to issue a complete account of all the assumptions that go into its black box.  All the projections about future catastrophes mean nothing if the assumptions of the PJM modeling are wrong.

As Mr. Herling points out in his PATH testimony, PJM only does transmission.  PJM has no control over the speed of implementation of DSM or new generation on the east coast.  Those alternatives are up to government regulators.  So far, those regulators, including the WV PSC, have failed to take responsibility for developing real alternatives to building bigger and bigger power lines that do little or nothing to solve our larger electrical grid problems.

Now is the time for the WV PSC to stand up and start working on solutions that go beyond PJM’s transmission system.