Skip to content

Could Microgrids Be Big in the US?

Three years ago this week, Hurricane Sandy began forming in the Caribbean Sea. It went on to wallop several Caribbean islands and the Eastern Seaboard. In the US the storm caused an estimated $75 billion in damage. The electricity system was hit especially hard. Over 8 million homes lost power.

Policymakers on the East Coast don’t want to experience Sandy-style outages again. That’s led them to microgrids, in turn popularizing the concept around the country.

So, what exactly is the benefit of a microgrid?

Enemies of the power grid gather. SOURCE: “Balloonsanimals” by RiseRover at English Wikipedia – Transferred from en.wikipedia to Commons.. Licensed under Public Domain via Commons.

To answer that, first we have to recognize that the electric distribution system is fragile. Harmless seeming trees, squirrels and Mylar balloons routinely cause widespread outages. Most grid incidents don’t just affect downstream customers, but also cause a ripple effect, knocking out upstream and adjacent circuits.

Other networks we deal with daily are more robust. If a tree falls on a road, traffic re-routes to adjacent roads. Global internet traffic is constantly rerouting around congestion and outages to get to its destination. Even a local internet problem can be mitigated by jumping on a cellular network or your neighbor’s wi-fi.

A microgrid, then, is intended to bring more resiliency to the electric grid. It ties together local distribution equipment and generators, and keeps them up and serving demand, even if the surrounding grid is down.

For example, San Diego Gas & Electric is developing a microgrid in Borrego Springs, California. The remote community is served by only one transmission line that runs through a fire-prone area. The microgrid, which combines energy storage and solar photovoltaics, is intended to keep the area energized when the transmission line goes down.

New Jersey’s public transit operator, NJ TRANSIT, is also investigating the feasibility of a microgrid to keep its rail system operating even when power outages hit certain parts of the grid.

The University of California, San Diego’s campus has developed a slightly different flavor of microgrid, one that goes beyond resiliency to market response. Consumption and generation on the campus can be optimized jointly in response to electricity prices. For example, the aggregate consumption can be reduced in response to high wholesale energy prices, or peaks can be shaved to avoid demand charges.

Additional companies and policymakers are considering whether investments into microgrids make sense. How should they assess the opportunity?

Reviewing the costs and benefits is a good place to start.

Costs include deploying distributed energy resources, installing advanced grid equipment such as automatic switches, and installing a microgrid controller to keep all the parts working harmoniously.

Distributed energy resources such as rooftop solar and energy storage face challenging economics, especially as regulators continue to reform retail pricing . Of course, that could change over time.

Meanwhile advanced grid equipment, like switches that automatically reconfigure the grid to avoid outages, are becoming more common and accepted by the industry.

Still, fundamental technological breakthroughs are also needed. The microgrid controller is a substantial challenge as explained in this Sandia National Lab report. The microgrid needs to recreate the functions of the bulk power grid in a highly local area. For example, microgrid supply and demand need to be kept in balance in real time. That’s tough on the bulk grid, but even more challenging on a small scale where there’s no room for error. Any entity pursuing a microgrid today will be taking on a research and development effort on these controllers.

On the benefits side, microgrids potentially offer improved reliability, lower costs and greater integration of distributed renewable energy.

I expect reliability is the most important benefit. Focusing on ways to improve electrical reliability in the US makes sense, especially given our poor reliability relative to other industrialized nations.

In a world full of microgrids, few customers would experience outages. Grid problems would be quickly isolated and distributed energy resources would supply customers who are cut off from the bulk grid.

However, utilities have many alternative tools available to enhance reliability. Burying overhead lines, building redundant connections and investing in automated switches are all tried-and-true solutions.

An honest evaluation of microgrids needs to consider these existing alternatives.

Microgrids could also lower costs in some instances. For example, a microgrid could manage peak demand served by a particular substation in order to avoid or delay an upgrade. Once again, alternatives need to be evaluated. The cost of upgrading the substation could be more or less than the cost of setting up and running the microgrid.

Using a microgrid to help integrate distributed renewable energy is another benefit supporters point to. The idea is that the intermittent generation from, say, distributed solar, could be matched to energy storage and consumer demand to keep a circuit from becoming overloaded. This approach would allow utilities to accommodate new distributed solar installations on a crowded circuit, instead of prohibiting them, as recently happened in Hawai’i triggering widespread protests.

Even faced with multiple hurdles and cheaper alternatives, investments in microgrids could still make sense.

Government R&D funders need to consider whether microgrids should be prioritized above other R&D objectives. I suspect that microgrids R&D would make the cut in the US given our sorry electrical reliability. We need to do something different on the grid. Microgrids could be part of the answer.

For private companies, the case is tougher. A company would need to believe that it can capture the R&D spillovers by patenting new technologies that are developed and selling them to others. That’s a risky proposition given the immaturity of microgrid technology.

In the near-term, the dominant model is government and private sector partnerships. Companies invest an amount that makes sense given their likely private gains, and government R&D pays the rest. That approach makes sense.

The microgrid of my childhood. SOURCE: By KMJ at de.wikipedia (Uploaded by Nordelch) [GFDL or CC-BY-SA-3.0], via Wikimedia Commons
Of course there’s also the “cool” factor.

Growing up in Houston, I remember getting excited as hurricanes approached. I enjoyed pulling out the candles and flashlights to prepare for the inevitable power outage. Perhaps the children of 2030 will get just as excited about the activation of their microgrid to get them through the storm.



Andrew G Campbell View All

Andrew Campbell is the Executive Director of the Energy Institute at Haas. Andy has worked in the energy industry for his entire professional career. Prior to coming to the University of California, Andy worked for energy efficiency and demand response company, Tendril, and grid management technology provider, Sentient Energy. He helped both companies navigate the complex energy regulatory environment and tailor their sales and marketing approaches to meet the utility industry’s needs. Previously, he was Senior Energy Advisor to Commissioner Rachelle Chong and Commissioner Nancy Ryan at the California Public Utilities Commission (CPUC). While at the CPUC Andy was the lead advisor in areas including demand response, rate design, grid modernization, and electric vehicles. Andy led successful efforts to develop and adopt policies on Smart Grid investment and data access, regulatory authority over electric vehicle charging, demand response, dynamic pricing for utilities and natural gas quality standards for liquefied natural gas. Andy has also worked in Citigroup’s Global Energy Group and as a reservoir engineer with ExxonMobil. Andy earned a Master in Public Policy from the Kennedy School of Government at Harvard University and bachelors degrees in chemical engineering and economics from Rice University.

12 thoughts on “Could Microgrids Be Big in the US? Leave a comment

  1. I’m a bit skeptical about the ability of a microgrid to deliver on its promises. For one thing, you’d need enough localized (or distributed) generation to supply all customers that are attached to the microgrid, and that’s not likely to be the case in most areas, (especially) urban or rural. For another, how would the microgrid operator (or Kristov and DeMartini’s DSO) compensate customer-owners of DG who sell their surpluses? Getting paid marginal cost is just not going to be enticing enough, especially if the DSO is the one who defines marginal cost and it uses the ridiculously complex rules ISO’s use.

    I’m told the UCSD microgrid failed to perform during the recent Southern California blackout that was caused by a lineman’s error near Palo Verde. If it’s true, what does that say about resiliency?

    In theory microgrids are a nice idea, but I’m not convinced they’re practical or worth the cost.

    • Jack, first technological innovation can be an amazing force. Two decades ago cell phone coverage was too unreliable to be used as a primary point of contact. Now many have abandoned their land lines. I expect we’ll see the same trend with microgrids and similar energy solutions.

      As for the market design, I don’t think that we fail to give enough consideration to the many different and wide ranging functions for transactions that can occur. Real estate and groceries are both markets, but have very different search and settlement processes. I wrote about this 20 years ago: We need to be much more expansive in looking for solutions that lead to appropriate compensation. Let’s put it this way: I don’t DSOs could do any worse than the CAISO is doing now in compensating true market competitors.

  2. The article points up the basic problem with today’s microgrid technology – while it would be valuable and many people would like the resiliency and other advantages of them, they are just way too expensive for the great majority of applications. Microgrids today are custom-designed and hand-built; technology that makes each installation an “R&D” effort is not scalable.
    Microgrids could be successful and highly useful if they were inexpensive by being built out of commodity components arranged into networks as needed. This is what we do for Internet Protocol communications, so how to do it is no mystery. Papers which elaborate on this are posted at under “Local Power Distribution”.

    • Bruce, looking at your listed papers, does a DC grid get around the need for inertia in an AC grid? Does solar perform better on a DC grid? (And what the heck is a “nanogrid”? My bedroom???)

  3. Micro-grids do have their place in the US. Aside from “Super-storm Sandy” NYC in particular demonstrates how well a networked distribution system can function. Even under Super-storm Sandy with Lower Manhattan out, upper Manhattan remained on.

    What do Manhattan, UC San Diego and Borrego Springs have in common? Density. Micro-girds work especially well in densely populated areas where distribution can either be placed underground or tightly networked together. The problem in the US is “space”. There are many places where there are long distances between load points making micro-grids economically unfeasible. Where it is much less costly to endure an outage than invest in high cost redundancy, storage and supply.

    Micro-grids will develop over time in densely populated areas. In places that are more “spread-out”, the economies of scale will need to be reduced to a point where the value of lost load exceeds the cost of investment in these systems.

    • I agree on your point about the likely locations of microgrids. However there are many dense urban opportunities. And microgrids can be strung together. I think this is the case for Borrengo Springs.

      A viable alternative for wide spaces has been spelled out by Kristov and De Martini as an independent distribution system operator who runs a grid below the substation voltage as a stand alone entity. It’s functionally not much different than a microgrid and eases the data management issues that an ISO faces.

  4. Few thoughts on this post:
    – Installing microgrids in “greenfield” developments is generally less costly than retrofits, and a utility can plan better for distribution investment.
    – The next question is who owns a microgrid? It could be a new income source for a project developer, but at least initially they would need to contract out for services. Or the governing jurisdiction could own and operate it along with other municipal utilities. Or the serving electric utility could own it. The answer affects…
    – R&D investment which might be better recovered by opening the market to competition. We don’t often ask how to recover R&D costs in other industries, because almost all of the players are “at-risk” in the market. It’s only a question here because utilities are so tightly regulated and aren’t allowed the variation in returns that come with R&E risk and opportunities.
    – Finally, there’s other benefits to microgrids other than just price and reliability. Microgrids may be easier to integrate with other types of resources and activities, such as using EVs as movable storage or using co-products within the building envelope. Given the proliferation of electricity management tools and resources, a centralized grid operator may have significant problems micromanaging the variations of integration. Lorenzo Kristov and Paul De Martini wrote about how the ISO may get overwhelmed by information and that little is lost in efficiency by disaggregating the arena to be optimized. In fact, given how CAISO grid operators have been overwhelmed at times with significant efficiency losses, moving to smaller decision units probably would increase efficiency. (Any study will need to look at actual dispatcher behavior and not rely solely on computer models.)

  5. Thanks for the posting. This comment solely addresses the question of whether the U.S. has poor reliability relative to other industrialized nations. The referenced basis for this is a prior posting that in turn relied on data from the Galvin Electricity Initiative.

    The Galvin data appears to be apples and oranges. Galvin cites to a European study which does not include the U.S. So Galvin’s U.S. data came from somewhere else. I suspect from a data source that includes “major” (weather) events. The European data generally does not include the rough equivalent called “exceptional” events.

    As I understand it if you take out the “major” events from U.S. data you are at 126 minutes of SAIDI and 1.08 of SAIFI — per slide 7 under “IEEE” columns here,

    This is much closer to the European results. Whatever difference remains after that could perhaps be attributable to other weather and population density.

    So I don’t think the data shows the U.S. to have relatively poor reliability.

    Thoughts on this comment would be most welcome. It is worth knowing whether the U.S. actually does have relatively poor reliability. Among other things a definitive yes would suggest we look at other nations for possible improvements, especially any “low-hanging fruit.”

    • Steve, Thanks for the comment. Even if the stats are adjusted as you suggest, the US is still #1 and over 5 times less reliable that Germany. More important than relative rankings would be an assessment of benefits and costs for specific reliability investments. We need to keep an eye on the impacts of major storms too. The realibility statistics typically exclude major storms to obtain an apples-to-apples comparison. Yet, storm-related outages are hugely damaging and could perhaps be reduced through grid investments. Maybe reliability statistics should be the subject of a future blog!

      • Andrew, thank you.

        I wholeheartedly agree that we need to do much, much more to assess benefits and costs for specific reliability investments. My article in the January 2015 Fortnightly on mandatory reliability standards urged that. The first part of the article is available here, and I’m happy to send the balance to whoever would like to read it.

        Back to the comparison of U.S. reliability with Europe, I’ve gone back to the source data, and the best I can determine:
        (1) the Galvin SAIDI/SAIFI table does seem to include all events for both the U.S. and Europe;
        (2) the difference in Europe from including and excluding “exceptional events” is small (comparing Tables COS 2.5-2.6 with Tables COS 2.1-2.2 of the European report referenced in my original comment) relative to the difference in the U.S. from including and excluding “major events” (see Table 4 of the Lawrence Berkeley report here,;
        (3) the Galvin table inexplicably excluded Finland, Norway, Portugal, and Sweden, all of which have much higher than average SAIDIs and SAIFIs (see Tables COS 2.5-2.6 of the European report referenced in my original comment);
        (4) considering all these European nations puts the U.S. in the middle of the pack for SAIFI and at the high end for SAIDI;
        (5) the U.S. being high for SAIDI might be attributable to the relative incidence of hurricanes and other extreme weather events in the U.S. that cause relatively long outages; and
        (6) the data also suggest that average length of transmission/distribution line per customer, with population density as a proxy for that, may partially explain differences in SAIDI and SAIFI (for example, at page 39 the European report finds that “… the continuity of supply improves when moving from rural to suburban to urban areas”); and
        (7) U.S. population density is a fraction of the density for all these European nations except Finland, Norway and Sweden (see

        Is this enough grist for the mill of a future blog?

%d bloggers like this: