A Pocket Full of PIMs

Michael Roberts, Blogs, Energy Policy and Planning Group, Energy, Governing Green Power

By Michael Roberts

Performance Incentive Mechanisms

In the arcane parlance of utility regulation, PIMs are “Performance Incentive Mechanisms.”

This is where we’re headed because, slightly against my expectation, Governor Ige recently signed SB 2939, a bill unanimously passed by the legislature that requires that the Public Utilities Commission:

“…establish performance incentives and penalty mechanisms that directly tie an electric utility revenues to that utility’s achievement on performance metrics and break the direct link between allowed revenues and investment levels.”

A nice summary can be found here. In the world of utility regulation, this act might be the equivalent of what Joe Biden called a “Big F[reaking] Deal” for health care. No other state has moved to change the investor-owned utility business model so radically, particularly to “break the direct link between allowed revenues and investment levels.” Like Obamacare, it’s still a compromise with the existing system, one that maintains our investor-owned utility but aims to change its incentives and business model in a profound way.

Just to be clear, PIMs are not new. They have been a part of utility regulation for a long time. But along the lines of what a colleague said the other day, existing PIMs are a bit like children splashing water against the hull of the Titanic in a futile effort to change its course. The core of the utility’s business model, in Hawai‘i and everywhere else, is revenue connected to an excessive rate of return on its own capital expenditure. This act promises to remove that, and make performance metrics the new profit engine.

Will it work?

Well, it depends. Like most things, the devil is in the details. It’s going to be important to get the performance metrics right, and to attach the right level of reward and penalty associated with each one.

In the old textbook model of utility regulation, the main performance metric is cost, and this is easily incentivized with a price cap, one that is just high enough for the utility to earn a fair return, and perhaps gradually lowered over time if the properly incentivized utility finds lower-cost ways of producing and delivering the goods. Sometimes the incentive to keep costs low needs to be buttressed with quality metrics, like customer service and reliability. This old Palgrave chapter by David Newberry has a nice review of the standard thinking.

But this standard thinking, or even Newberry’s longer treatment, probably won’t work with electric utility of the future currently envisioned in Hawai‘i and other places embracing renewables. What’s new is the growth of distributed resources. This isn’t just rooftop solar. It’s batteries, hot water heaters, air conditioners, electric cars and all manner of electricity uses that have potential flexibility in their timing, and can therefore be employed in a way that makes management of intermittent renewables less costly. This changes standard regulatory frameworks because the utility is unlikely to own most of these flexible distributed resources, yet the way they are used and integrated into the system is key to keeping costs low. And since these resources can compete with the utility’s own investments, it creates a palpable conflict with existing assets owned by HECO, those seeking to sell renewable energy storage services, and with customers.

In other words, the integrated system’s costs no longer equal the utility’s costs.

This new law directs the Public Utilities Commission to reconcile conflicts by using performance metrics to better align interests and find the least cost path toward a renewable energy future.

All of which begs the question: What are the best metrics?

So far, the proposed metrics target plethora of narrow issues, hence the title of this post. These fall well short of an encompassing framework that could redefine the utility’s business model. But I think it’s possible to build such a metric. Economics provides some guidance, with reasonably comprehensive monetary metrics of the net social value of our electricity system, including distributed resources. We could even add costs of pollution externalities to such a metric.

Some may quibble with some of the assumptions, and that’s an important conversation to have and revisit regularly. But if a reasonable consensus can be achieved about both the assumptions and the model, the PUC could tie the utility’s allowed revenue to a measure of this net social value. Or, more precisely, the difference between a reasonable, agreed-upon target for this metric and the outcome actually achieved. The target would be tied to clearly identifiable input costs, like fuel prices, the cost of renewable energy, battery storage costs, and even costs of capital, so that performance would hinge on how well the various resources are integrated and managed, if the utility can negotiate good deals for certain inputs, or avoid unnecessary expenditures through smart management. Utility ownership of assets would have nothing to do with it — the utility would simply make more money if it facilitated more social value.

This idea bears more than a little resemblance to a standard way the PUC has always managed rate cases every three years or so to set allowed revenues and rate schedules. When they do this, the utility, PUC and consumer advocate rely on optimization models that are used to estimate costs and set allowed revenue. The idea laid out above bears some similarity to this process, except that it would encompass costs and benefits that would in one way or another extend beyond the utility’s own costs. The new big piece is incorporation of the demand side–-customers’ benefits from electricity use. That’s something myself, Mathias Fripp and Imelda, a UH Mānoa PhD student, recently figured out how to do in order to show how much variable pricing and demand response can lower the cost renewable energy.

In conventional utility regulation, the model is used as a cost baseline to set allowed revenue. Here it would set the target level of social value.

Allowed Revenue = a + b ( social value – target social value)

Besides the model assumptions, which would set the target social value, the PUC would need to set levels for a and b, which set baseline revenue for the utility and the degree to which the utility would be rewarded or punished for exceeding or falling short of the target.

All of this may seem abstract, and on some level, it is. I can’t argue with the idea that this is a “black box” model (except that that is publicly available, and there is always a black box model). It will take some work to make it clearer for a broader audience. We’ll work on that. It’s still early….

Stepping back, the big picture idea here is to cut all engines on the Titanic and harness the coordinated efforts of its captain together with those of a hundred tugboats to push the ship in better direction for everyone. The metaphorical iceberg here is mass grid defection, which I fear could be more likely, and sooner, than many realize.