Tuesday, November 03, 2015

Another network you’ve joined

My post earlier this week about incorporating end users of the Internet into the network to enhance its performance and stability reminded me about a research project I was involved in back in 1979 at MIT.A group of us, led by the late MIT professor Fred C. Schweppe (one of the world’s experts in electric power systems control) took a look at how a dynamic electric power system might function.  The term for the project was coined by project member Richard Tabors, who was trained in biology: “Homeostatic utility control.”

As Richard reminded us, a human body responds in real time to changes in the environment and other challenges to moderate heart rate, respiration, and the like to keep us on a steady keel. We eat that Halloween candy, and our pancreas figures out how much insulin we need to convert sugars. We take a run to work off the candy calories, we get hot from the exercise, and we sweat to cool down our bodies.  We face the danger of a territorial dog as we run, our adrenaline flows, and the hormone stimulates all kinds of physiological adjustments, which then relapse to base levels when the threat is gone.

We all wondered whether a regional electric grid could do likewise if signals to and from the periphery were employed.

Specifically, the question we sought to answer was whether an electric power system that incorporated demand side management and disbursed power generation at the site of customers would, with a pricing regime that transmitted the real-time marginal cost of electric power production, be stable.  In other words, would customer responsiveness to actual minute-by-minute prices result in a more efficient system—or would it spin out of control in a fit of instability. (Remember, in a power grid the demand for electricity has to be supplied essentially instantaneously to maintain voltage levels, frequency, and other key parameters.)

Our research question was hypothetical at the time.  There were very few distributed electric power sources in 1979, and what energy conservation and load management programs existed were not premised on customer response to real-time price signals.  But we all envisioned a world when such things would be commonplace.

Well, the analysis showed, as well as could be determined based on lots of assumptions, that a network that integrated customer-level supply and demand price-responsiveness could indeed be stable (as stability would be defined by power systems criteria.)

We fast-forward to today, a world in which distributed generation and load management are rapidly infusing regional power grids.  Think about those solar panels on your roof that supply you and also sell power back into the grid, or appliances that respond to price signals to operate during off-peak periods. With improvements in computer technology and telecommunications infrastructure, it is likely that the conclusions about stability we reached back in 1979 are still true.  Today’s power systems, too, are likely to have more resiliency than the power systems that existed back then.  There is less dependence on a few, very large generating units.  There is more redundancy in electricity transmission capacity.  And locating power generation closer to load centers reduces a portion of transmission system losses that occur when power is sent over long distances.

I’m currently on the board of one of the organizations with some responsibility in this arena, ISO-New England, the supervisor of electric system reliability and the entity that designs and operates the markets for capacity and energy in this six-state area. You can imagine that the many sectors in this regional power exchange (e.g., transmission owners, generation owners, investors in “alternative” resources, end users, power marketers, and others) have divergent financial interests. I’ve been tremendously impressed, during my tenure on this board, at the degree to which the participants work together—notwithstanding their individual financial interests—to help design a system that works for the benefit of the consumers in the region, with due consideration for environmental protection.

But good will and a sense of regional purpose will not necessarily yield consensus in this forum or in similar forums in other regions of the country.  Among other things, participants have different time frames over which they engage in capital formation, cost recovery, and return of capital.  Important public policy questions remain, too, because the electric power system is not only ubiquitous, but because its output has become an essential input to our lives—affecting public safety, education, other utility services, our personal lives, and commerce.  Thus, the manner in which distributed generation and load management are introduced into the electric power system will remain a topic of hot discussion in the body politic and before federal and state regulators.

Contrast this vibrant debate with my earlier post, in which I noted that Akamai’s reach into the distributed network to be present on your home computer occurred without much public discussion or government regulation. I do not argue with the technical merits of the solution employed, but I note that the question of how the resulting enhanced value of the Web is shared among its participants has not reached the public consciousness. Maybe things are going so fast because of the exponential growth in Internet traffic that such questions can only be viewed through a rear-view mirror.

While the stakes for the Internet are high, even the Web requires electricity to run. I urge my readers, whether running hospitals or involved in other industries, to become informed as to the issues facing our electric power system and not only be wise buyers and sellers in that marketplace but also active participants in the public discussions surrounding this sector. Our job, together, is to help ensure that the rules governing the enhanced distributed network do not produce a zero sum result but rather create value for society as a whole—and that the manner in which that value is shared across the sectors is broadly viewed as fair and proper.

--
* Summarized here: "New Electric Utility Management and Control Systems," MIT Energy Laboratory Technical Report, No. MIT-EL-79-024.

No comments: