The weird and unique creature of electricity markets


“What the Hell is this?”, my CEO said as he was looking at a pro formafor financing a major acquisition. Twenty or so years ago, my company (US Generating Company – it would become part of PG&E National Energy Group) was bidding on the generating assets of New England Electric. My CEO was Joe Kearney, the best leader and smartest man I ever knew in the industry. A PhD graduate in nuclear engineering from MIT and a former developer of natural gas pipelines, he was also as facile as anyone I ever knew with finance… you know, how we get money to fund things.

What Joe was referring to in his question was the estimated “value” of the “capacity” of the assets we would ultimately acquire. The essence of capacity – being paid for being there, was explained to him. He was incredulous. “You mean, they are going to set up a clearing price for just being there?” Yes, he was told, it was a matter of reliability – especially as electricity is not easily or (even today) inexpensively stored. Indeed, as I sought to explain the capacity concept to a banker years later, the best description for me was as a “proxy for storage”.

In our current organized electric market construct “capacity” is valued in many different ways. In three Eastern RTOs, there are varied kinds of “clearing markets” – some multi-year, some annually. These are often called “markets” but they really are “mechanisms” intended to ape a market response and have high levels of administrative assumptions built into them. In SPP and MISO, there is no centralized administrative mechanism for clearing capacity – although MISO runs an auction to help with price discovery. ERCOT, of course, runs the closest thing to an “energy only” market – with its obligation on the load-serving entities to procure enough capacity to meet reliability and customer needs. All of these markets have States which have policies for renewable generation procurement by utilities – but there are no methods to price these all of these resources differently from one another except on the basis of availability.

All of these markets – despite their differences – have a single mechanism for attaining an adequate amount of resources to serve customers (or to use the ugly electric term, “load”). The sole exception to this unified approach to “resource adequacy” is California. The CAISO has a requirement that entities that serve customers (OK, “load serving entities”) procure an appropriate level of resources to meet reliability. This is essentially a “bilateral” market with willing buyers and sellers. Fine, as far as it goes. Then there is the pesky need to sometimes procure a resource in an area that needs electricity in a hard to serve area (could be land restrictions, transmission limits, etc). These are intended to be procured efficiently but sometimes there is only one real alternative and terms are agreed between the CAISO and a resource owner. The point is that the entity that is responsible for reliability – the CAISO – is supposed to make the call.

If this were the case, things might work out very well. What about pushing for more renewable assets or battery storage, you say? Well, a robust cap and trade market there is a real incentive for these resources to be procured. Additionally, California is in the midst of a boom of customer aggregation from investor-owned utilities to “Community Choice Aggregation (CCA) providers. Many of these new efforts to serve the customers of particular communities are done specifically to procure more “green” assets. With the explosive growth of the CCAs one can easily imagine more and more demand for non-thermal (natural gas, et al) assets. But this is where rational behavior ends and arbitrary decisions beings: the California Public Utility Commission( CPUC) likes to get in and tell the utilities it regulates to not to do business with those dirty natural gas plants but rather do business with “preferred resources” like renewables and storage.

There are two huge problems with this bi-furcated and inconsistent approach. Judging from recent decisions of the CPUC that have ordered a re-opening of agreements on agreed reliability procurement, the demand for “preferred resources” is made without reference to whether these assets can meet the amount and duration of the reliability need. This is problematic because the sun doesn’t always shine, the wind doesn’t always blow and batteries can’t discharge power at consistent levels for long durations. The second problem goes to ultimate fairness and economic viability. Let us take as a given that thermal assets are needed for the system as no rational person suggests otherwise in the near term. Accepting this inconvenient fact: how does one expect the continuance of these resources when the CPUC – or other agencies – can step in and change the decision of the CAISO or the servers of load to contract for to meet their needs? While nobody would suggest that these assets must be given handouts to survive, they do need a rational  system for contracting.

As someone who was at PJM when its current capacity construct was created, I am less and less of a fan or centralized capacity markets of multi-year duration. There are too many administrative elements to it – which are often based on subjective elements -for such mechanisms to achieve the efficiency of the market. But is the multi-layered programs being pursued in California ensuring reliability over the long term? State regulatory Commissions are under a great deal of pressure to give wing to the whims of politicians. Nevertheless, I hope that we can engage both the CPUC and the CAISO to review how they procure resources so that reliability is assured, while recognizing there is no need for such a review to compromise the other policy goals of California.