GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Guest Blogs

Why We Still Need to Discuss Grid Defection

This relatively new topic for the utility industry should prompt a careful look at the research

Rapidly falling costs for distributed energy, including rooftop solar, are introducing new dynamics into a traditionally slow-moving electricity industry.
Image Credit: Jon Callas / CC BY 2.0 / Flickr


The rapidly declining costs of distributed energy resources (DERs), including rooftop photovoltaics (PV) and behind-the-meter batteries, have introduced new dynamics into a traditionally slow-moving electricity industry. This paradigm shift has ushered us into a new era where previous assumptions about how, where, and at what scale electricity is best generated, transmitted, and distributed may no longer hold.

In 2014, Rocky Mountain Institute (RMI) released a groundbreaking analysis of the potential for “grid defection” — when and where it may be economical for customers to disconnect from their utility in favor of using on-site solar + battery systems. In the years since, as DER technologies have evolved and costs have come down even further, a number of fellow analysts have offered their own interpretation of the economics of these technologies and the implications for electricity industry stakeholders.

For example, a recent study by Eric Hittinger and Jawad Siddiqui for the Rochester Institute of Technology (RIT) reexamined a subset of the issues RMI addressed in our original 2014 paper. Although Hittinger and Siddiqui reinforce many of the conclusions we made in the paper, media coverage of the RIT study contends that it “throws cold water” on the economics of customers using solar and storage to defect from the grid, and in particular RMI’s 2014 findings.

We acknowledge that because grid defection is a relatively new topic for the industry, we must proceed with humility and openness to contrasting views; however, we also believe that it’s important to clarify misconceptions regarding the research RMI has conducted to date on this topic and its consequences for the electricity ecosystem.

A narrower scope limits the interpretation of new results

The RIT study takes the analytical structure of RMI’s 2014 paper and applies it to more locations in the U.S., with more-specific technology options. It arrives at many of the same conclusions as our 2014 study did for the present-day potential of grid defection (namely that grid defection makes economic sense today only in Hawaii where retail electricity rates are the highest in the nation).

However, it also makes several assumptions that differ from ours. First, the study only examines current DER costs without considering future trends, as our 2014 paper did. Second, it does not consider the full suite of options available to customers, as our more recent studies — The Economics of Demand Flexibility and The Economics of Battery Energy Storage — explore in greater detail. Finally, it misses the point of why the future potential for cost-effective grid defection matters, even if most customers may not ever consider it.

(1) Focus on present-day costs. The RIT study, which covers present-day PV, battery, and grid costs, does not consider how these costs may change a decade from now (i.e., cost declines in solar and storage, and cost increases in grid-supplied power). RMI’s original study explicitly aimed to highlight these points, studying cost trajectories from 2014–2050. Without acknowledging the future potential of cost-effective defection, the RIT paper misses the implications that a big portion of the ~$100 billion per year of utility capital investment today could be stranded by future customer investment. Major investment banks, among others, have also seen this potential danger to the current utility business model if present trends continue.

(2) No consideration for a broader suite of DER options and use cases. The paper sets up a comparison of customer costs under a limited set of solar power scenarios (e.g., net-metered PV versus complete grid defection), but ignores other DER options that are potentially more likely. The paper does not consider options we explored in RMI’s more recent work that assesses the present-day customer value of demand flexibility under real-world utility rate structures, or the potential to gain expanded value from behind-the-meter battery energy storage systems. We have found that these middle-ground solutions, stopping short of full grid defection, offer much better economics for the customer, and likely for the utility and larger grid as well.

(3) Why grid defection still matters. By failing to consider both the implications of changing future costs of defection, and the present-day cost-effectiveness of other DER solutions, the RIT paper misses an opportunity to emphasize why grid defection matters today even if it is only economic in the future — namely, that it empowers customers to determine the pace of decarbonization, retail pricing evolution, and other grid investments (e.g., in resiliency) that would otherwise be left entirely up to incumbent utilities.

Because of these differences, the RIT study doesn’t actually challenge the core conclusion of our 2014 study and our more recent papers: that the rapidly evolving costs of the full range of modern DER options make distributed, customer-centric solar power increasingly practical and competitive in a rapidly expanding range of conditions. We believe this conclusion is more important and indicative of the future than absolute grid defection would be.

Misinterpreting the results

Because of the framing of the RIT paper, it may reinforce some mistaken beliefs on the evolution of the electricity grid that are worth noting.

Claim 1: The “grid as a battery” for customer PV is the most economical choice. Hittinger and Siddiqui’s comparison of net-metered PV versus grid defection may suggest that the former option, net-metered PV, is always the least-cost solution. Yet both the RIT study and RMI’s own recent work have laid out that while net-metered PV may be the most favorable outcome for individual customers today, there are likely better ways to assess the value of and remunerate solar PV and other DERs on the grid. “Using the grid as a battery,” with solar PV and no other DERs like efficiency or demand flexibility, is likely to require continued investment in expensive and lightly utilized assets that utilities will find difficult to justify. It is important to encourage customers to manage some of their own capacity and flexibility needs, for example through updated retail pricing.

Claim 2: Utility ownership of PV and other DERs is always the right solution. It is easy to take the negative results of the RIT study for the specific customer-sited DERs under consideration and assert that utility-scale and/or -owned DERs must be the most cost-effective choice. However, RMI’s recent work on the value of flexibility and storage has made clear that utilities, instead of owning or directly deploying DERs, can also reap their benefits by passing the correct price signals to consumers and/or aggregators. California, New York, and other leading markets are actively exploring these approaches.

Customers will play an important role in the integrated grid

A conclusion that grid defection is not cost-effective now in most of the country glosses over the other ways in which customer preferences and investment decisions are actively transforming today’s grid. In our aforementioned papers, we detail how customers can invest cost-effectively in DERs (e.g., smart thermostats, batteries, electric vehicles) that provide better services and can reduce grid costs at the system level. Although these technologies do not amount to grid defection, they do indicate that DERs are playing an increasingly important role in the architecture of the grid.

We continue to believe that increasingly cost-effective DER options will encourage customers to take a more active role in their energy, and lead utilities more rapidly into a new kind of electricity system, as we said in one of our earliest blog posts. Whether that leads to customers defecting from the grid or not, these customer choices will continue to transform the grid outside the scope of utility planning processes and regulatory structures. The lowest-cost and most desirable outcome for all customers will embrace this dynamic and seek to keep these customer-sited technologies integrated with the grid, rather than apart from it.

RMI recognizes the systemic challenges presented by widespread grid defection, and will continue to advocate for an integrated grid. The “fork in the road” message we used in our 2015 follow-on study to the grid defection work, The Economics of Load Defection, frames how reforms in electricity tariffs, utility business models, and utility regulations are crucial to arrive at an integrated, resilient, and cost-optimal grid. RMI’s ongoing work in the electricity sector is tightly focused on that outcome.

James Mandel, Ph.D., is a principal at Rocky Mountain Institute. Mark Dyson is a manager with RMI’s electricity practice. Todd Zeranski is a marketing manager at RMI. © 2017 Rocky Mountain Institute. Published with permission. Originally posted on RMI Outlet.


  1. SwitchgrassFarmer | | #1

    Haves vs Have Nots, Suburban vs Urban & Rural
    I worry about the social implications of "grid defection", particularly in light of an increasing divide between the "Haves" and "Have Nots" of the world.

    On the surface it would seem that grid defections will ultimately drive up the cost for those that can't afford their own generating and storage system. For example this morning I read this piece in the Washington Post: Sadly this reflects the situation in wide swaths of America. How do these "grid defections" not inevitably raise the costs for folks in such dire straits? No doubt this is an argument that utilities make.

    It also occurs to me that "grid defection" may be a somewhat suburban centric concept; it works for homes with solar panels on the roof. Large sectors of the population live in dense urban centers where that isn't possible, so they may be penalized too. BTW, those urban centers, perhaps even better on an overall environmental perspective due to the availability of other common infrastructure like mass transit, water systems, etc.

    Anyhow those were the first thoughts that went through my mind after reading this article; I recognize this is a simplistic knee jerk reflex. I look forward to what others who have studied this issue in depth (Dana?) might have to say about it.


  2. GBA Editor
    Martin Holladay | | #2

    Response to Andrew Bater
    I think that your worries are legitimate. You are correct that suburban and rural households are more likely to defect from the grid than urban households, and that wealthy Americans are more likely to defect from the grid than poor Americans.

    That said, the trends leading to increases in grid defection are likely to play out, whether the results are socially beneficial or socially disruptive. Those of us who care about social justice will simply need to redouble our efforts to create a more just society for all.

    -- Martin Holladay

  3. JC72 | | #3

    My two cents,

    I suspect government will levy a "grid tax" on all grid defectors to help pay for grid maintenance.

    On a side note I read the linked story. I feel for the guy who kneecapped his life by becoming a felon (twice) before age 20. Very few employers can take on the liability of hiring such an individual.

  4. itserich | | #4

    more time than money
    Air sealing and insulating is a great activity for people with more time than money. Growing food is also a healthy activity that can be done cheaply, or not.

    It is hard to imagine during what time period enough homes will go off the utility grid to be noticeable. It would also be interesting to figure out a rationale and a way to tax those households. Homes which use the grid as a battery should pay for the battery, but going off the grid is completely different.

  5. vensonata | | #5

    New economics of energy independence.
    "Grid defection" is already a loaded term. Try "energy independence" instead. American national energy policy does not use the terms "World energy system defection" they use the phrase "American energy independence". Even the right wingers like the idea because it keeps them "out of wars". And in fact, the individual can also end up in a conflict with "big grid", who don't necessarily have anything but the profit motive in mind.
    Anyway, the economics of energy independence have changed, it is now, or very soon to be, affordable and convenient. It is usually the math that boggles the other words, how to make a simple picture of big grid vs micro grid.
    The other part of the picture is the electric car. That changes everything. The combined energy costs for a household plus gasoline on average are $5500 per year. Multiply by 30 years and you will spend $165,000 without considering rate hikes etc. What kind of system can you purchase that could provide 30 years of energy for house and vehicle that costs less than $500 month of loaned money? A very large system of PV and storage is the answer. To keep this short, I won't provide detailed cost and production, (but see what you can buy for say, $80,000) but it is definitely reasonable as a competitor, if you have a house with a decent solar angle. There are probably 30 million of these in the U.S.

  6. Expert Member
    Dana Dorsett | | #6

    Grid defection is a real problem
    If the evolution of regulations gets to the point where out and out grid defection is the best course of action for those with the means to do so, it's a real problem. Those assets have far greater value to others connected to the grid if they stay hooked up. Defection also pushes the grid infrastructure costs onto the dwindling number of people with no option to leave, and it becomes unaffordable. Getting to that point would be (or rather is now, in some locations) a failure of regulation & policy.

    This isn't just theoretical. In Nevada ( where policy & regulation mismanagement has been a total zoo) there have been a couple of high profile grid defectors in recent years eg:

    If it's cost effective to even pay the exit fees to quit the grid, smart money managers with sufficiently deep pockets will do that!

    There are locations in Australia where high fixed connection fees and low remuneration for exported power to the grid from PV owners have made it financially rational to quit the grid too, and unless the utilities (and government) figure out more equitable ways of dealing with it, and actually UTILIZING the grid-stabilizing potential of behind the meter assets, there's a real risk of a "utility death spiral" getting started. The costs of battery technology are coming down faster than anyone was predicting even 2 years ago, and it's quite scalable to the single house level.

    It takes a bigger PV array and more battery to quit the grid than it does to be net-zero-electricity while still hooked up to the grid, but not vastly bigger- it's not even 2x. But the overall capital expenditure by utilities + homeowners is quite a bit less if they stay connected. Policies that promote staying connected and sharing the resource are important for keeping overall costs bounded, and keeping electricity affordable. That is a major factor at the heart of New York's major overhaul of utility regulations in the Reforming Energy Vision, which is still a work in progress, but already saving money:

  7. itserich | | #7

    MGM still receives a majority of energy through grid
    MGM is still purchasing a majority of its energy from other producers, and that energy is provided through the grid.

  8. iLikeDirt | | #8

    I plan to do it
    I plan to do it myself, for Ven's reason: independence. The economics are racing in the direction of it making sense, but it's not selfish or mooching off others to be independent. On the contrary, it increases society's resilience for more people to be able to provide for themselves when the big systems get disrupted. This is not at all a left-right issue, if indeed it is political at all. Both parties have factions that prefer interdependence and factions that prefer independence..

  9. JC72 | | #9

    @Nate G
    Makes sense if people can just do it without taxpayer support.

  10. exeric | | #10

    @ John Clark
    "Makes sense if people can just do it without taxpayer support."

    Totally agree. I also think that its time for people who can afford it to build their own roads locally in their community. What right do poorer people have to milk the government for roads and infrastructure that wealthy individuals can build and create for their own individual use? And don't get me started about medicare and social security. It's high time people quit mooching off the government and started being responsible for their own lives. It may be considered politically incorrect by some sensitive types but perhaps we should return to days gone past. Was it really so bad in medieval times with vast numbers of serfs serving just a few lords and ladies. I think not.

  11. Expert Member
    Dana Dorsett | | #11

    A glimpse of the future (with the right policy support)

    If the utilities & regulators make staying hooked up too expensive nearly-net-zero and net-zero homes will unplug completely, but it's not as easy (or cheap) as it might look at first blush:

    But it's getting cheaper every day.

  12. GBA Editor
    Martin Holladay | | #12

    Response to Dana Dorsett
    Thanks for the links. For off-grid veterans like myself, I just shake my head at the naïveté of the newbies who jump into the off-grid lifestyle. They should talk to a few people who have been living off grid for years and listen for a while.

    None of these lessons are new. There seem to be lots of people who are excited at the idea of going off-grid, but not willing to do even a little homework.

    -- Martin Holladay

  13. exeric | | #13

    @ Dana
    That second link you provided is exactly the reason that battery storage is never going to be the ultimate solution, just part of the solution. The real answer is battery storage plus the idea Bronwyn's blog brought to everyone's attention months ago. That is, hydrogen generation, or better yet methane production through techniques related to electrolysis. Of course the people fixated on their own energy independence will never like that because the economics of that will always require a centralized generation and storage facility that utilities run. Of course if some rich person, or persons, can own it and collect rent instead of a utility then all would be fine, because then the government and regulations would not be involved.

    There will always be periods of extended overcast for nearly all geographic locations that will make batteries only an incomplete answer for individual energy independence. Maybe parched parts of Australia and the Sahara and Mojave deserts are an exception.

  14. Expert Member
    Dana Dorsett | | #14

    Staying on-grid (however lightly used) is the ultimate solution.
    At the current price of storage you can build out 2-3x the amount of solar needed to hit net zero rather than over-building storage or expensive seasonal storage to deal with variable seasonal output. That would involve a lot of curtailing of output, but when local PV hits 50cents/watt, installed the cost of the electricity is low, and the cost of curtailment is also low- it simply won't matter. But that's not the optimal expenditure of capital.

    But staying hooked up to a grid, with a diversity of generation types and averaging power over large geographic areas will make sense for the forseeable future, even as behind the meter generation & storage proliferates. The difficulty is how to get there in a manner that doesn't screw the non-generating ratepayers or sink the utility/grid operators, and it involves major revisions in utility business models to make it equitable. The New York REV is an attempt to get out ahead of it rather than end up in the crazy situation much of Australia is currently dealing with, where utilities and grid asset owners were promised a return on sunk infrastructure costs that are now inappropriate and underutilized, and the current & change regulations often stand in the way of doing the right thing.

    Net Zero Energy will be the law of the land in California for new residential construction in only three years, but nearly-net-zero-electricity will become economically viable for many existing homes in under 10 years. That makes financing large central generators and many grid assets a much riskier proposition than it was in the 20th century. But being able to run the heat pump for a Net Zero house in Ohio that has a 6kw array with offshore wind in New Jersey is probably going to make a lot more sense than building out 20kw of PV and a battery big enough to cover worst case scenarios, and curtailing 2/3 of the output of the PV for at least then next couple of decades.

    Grids are good, but the business models and regulations have to evolve to maximize the value before too many people cut the cord. My best guess (and it IS just a guess) is that they have until 2030 to figure it out, but it could be much sooner. NY will get there in time, and TX probably will too, but GA and other states with large vertically integrated monopoly utilities might not.

    The whole seasonal storage thing is a bit overwrought. Hydrogen & methane generation & storage isn't free, and the costs have to be less than curtailment of PV to be economically rational going forward. The sun really is always shining somewhere, and the wind really is always blowing somewhere. The amount of short term or seasonal storage needed is much smaller than most people imagine, as long as we don't all pull a Michael Mobbs and try run our own private nano-grids isolated from cheap & green power that might be available 1000 miles away, or just down the street. Yes, there is already a substantial amount of pre-existing gas storage and pipeline assets, but it'll still come down to the economics of generating & storing gas fuels compared to solar that has been and still is on an unrelenting & steep learning curve. Widely distributed PV and small amounts of storage can free up an amazing amount of capacity on the existing power grid, which allows greater use of that grid to move energy over large geographical areas to meet peak loads.

  15. exeric | | #15

    @ Dana
    Dana, I agree with most of what you said. I didn't say that non-battery local storage need be a high proportion of total infrastructure. It all comes down to minimizing capital expenditure to use all the buzzwords. I'm no expert on the proportions of each where that minimization would lie. But I totally agree that the existing grid is a huge asset that we ignore at our peril. And the money sunk in that is something we should all value just like we value roads and clean water. There is way too much fetishism of rugged individualism in this country where people become deluded that they are the only ones not feeding off the government teat. Everyone does to a greater or lesser extent. And it's not a bad thing. It just has to be managed in as fair and equitable manner as possible. Nothing is perfect.

  16. JC72 | | #16

    The sooner you understand the coercive nature of The State the sooner you'll realize that there's no such thing a free lunch and it is the poor who end up paying most of the tab.

  17. Expert Member
    Dana Dorsett | | #17

    Meanwhile, in the land of Oz...
    Meanwhile, the grid operator in Australia is projecting that by 2035 around 70% of all electricity customers will have behind the meter PV, and 50% of all customers will have behind the meter storage:

    If the regulators don't screw it up, maybe only a few will actually cut the cord.

    In the US something like half of annual power COULD be sourced from PV on existing rooftops (not that it ever will):

    The size of the resource is large, and the amount of new real estate required to deliver a very large fraction of US power (in any season) is small.

    As long as utility regulators in the US don't screw it up, not too many will be cutting the cord here either. Cutting the cord will still be pretty expensive for the forseeable future.

  18. EthanT | | #18

    make sure you trade your prius for an suv too...
    ... you wouldn't want to be skimping on gas taxes...

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |