The Problem with 100% Renewable Energy

The New York Times reported recently on the work of a group of Stanford researchers, who published a piece in the journal Energy Policy claiming that the main issues with 100% renewables are neither economic nor technical but rather social and political will. Jesse Berst of Smart Grid News countered with a title that included the phrase, “Get real, Stanford.” As experts in the economics of renewable energy, HOMER Energy would like to explain why we think that the promise of 100% renewable energy as a purely political problem detracts from getting where we need to be – which is at much higher levels of renewables than currently. Ignoring the very real economic issues (and opportunities) that exist with renewables will not get us there.

The HOMER software models the lifetime costs of power systems that include combinations of energy sources – from purely conventional generation to 100% renewable. HOMER is an accepted standard for optimizing economics of hybrid systems that has been used throughout the world.

We used HOMER to model the  Levelized Cost of Energy for renewable penetrations from 0-100%, on an island with a diesel-powered grid, representative of hundreds of power systems that we have modeled at HOMER Energy. Levelized cost of energy (LCOE) is a way of creating an “apples to apples” comparison of energy from renewables, which is often front-loaded with capital costs, to energy from diesel generation, which has low capital costs but high operating costs.

The renewable penetration is represented on the horizontal axis of the graph as a percentage, and LCOE is represented on vertical axis, in dollars per kilowatt-hour. 0% renewable penetration means the island is completely dependent upon diesel generation for its electricity.

Cost of renewable energy as a function of RE penetration

The first thing to notice is that at low levels of renewable contribution, renewable generation reduces the cost of power. This cost reduction results from displacing expensive diesel fuel with low-cost intermittent renewable energy, such as wind and solar (solar photovoltaics reached price parity with diesel fuel in 2011).

Adding more renewables beyond  about 40% requires investments in integration controls. This cost, combined with the occasional need to curtail some of the renewable generation, reduces the economic attractiveness of additional renewable generation, but still allows very substantial levels of renewable contribution, with their additional environmental benefits, with only modest cost impacts.

When renewable penetration exceeds  about 80%, however, the cost of meeting the electrical load goes up substantially. This results in the “hockey stick” price increase and makes it prohibitive to achieve 100% renewable contribution without some form of liquid (i.e. easily storable) renewable fuel.

To summarize:

  • At low penetrations (0% to 30%), simple substitution of diesel energy with renewable energy yields cost reductions.
  • In the middle region (30% to 80%), the cost of integration technologies flattens the curve.
  • At the high end (80%+), the requirements for excess renewable energy capacity coupled with very large storage requirements drive up the cost of energy very steeply.

The implications of this shape are enormous. One the one hand, it establishes an economic case for integrating not just low-contributions of renewable energy into these grids, but medium, and even high, contributions – up to around 80%. On the other hand, it also demonstrates the prohibitive cost of exceeding the 80% threshold with current technologies.

Many systems could be brought to 70-80% renewable for the cost of taking one system to 100%.

Political or social will are important, but they can be undermined by promoting excessively expensive solutions when much greater benefits can be achieved more quickly and broadly by promoting more cost-effective solutions.

Dr. Peter Lilienthal, CEO, HOMER Energy
John Glassmire, Senior Energy Engineer, HOMER Energy
Dr. Marilyn Walker, COO, HOMER Energy


    1. Hi Mohamad,
      You are very right, and thank you for your comment. There are a lot of topics to discuss around 100% renewable energy as the technology changes, and as governments and companies pledge to use cleaner energy. We are working on an update.
      Lili Francklyn

  1. sorry, just see this has been discussed above – in this case i find the right end of the graph rather misleading.

    1. Yes, biofuels would reduce the steepness of the right side of the graph. I think the basic shape would be the same, but less extreme. I also worry that biofuels are more valuable for transportation, so it depends on how much biofuel production is possible, which is outside my expertise. It also depends on how well EVs penetrate the market. They would not only reduce the demand for biofuels, but also provide a large and flexible load to the utility grid. There is a lot that can be done to reduce the steepness of the right side of the graph with more flexible load management. I still stand by the fundamental point that I was trying to make, namely that focusing on the last 20% is a distraction, because so much more can be accomplished by focusing on what is achievable at reasonable cost now.

  2. just stumbled across this nice graph – and wonder if this considers that you can just use biofuels in backup-generators for the last % – i believe the last 10% do not need to be so expensive then?

  3. Have you read the paper recording the simulations of 90+ renewable power done at the University of Delaware and published a few months ago?

    Their model says that very little storage is required … just a variety of renewable capacity at a variety of locations.

    1. No, I hadn’t heard of that study at the University of Delaware study. Was that with Dr. John Bryne’s group that shared the Nobel prize with Al Gore and the UN ICT Team? It will take a combination of fuel cell, battery, pumped storage hydroelectric, biomass fuel from digesters, feedlots, landfills, abandoned coal seam beds, alcohol from swatch grass and sugar beets, wind, tidal, river currents, photovoltaic, mass transit, passive solar heat, biodiesel and geothermal energy to meet our energy needs. It will depend on a “smart grid”, the geography and access to raw materials. It will be both centralized and decentralized, depending on the geographical location, size of the buildings, businesses and industries involved.

      1. Hi Lyman,
        We have responded to your question via email. Many thanks for commenting.

        1. Please how can we get this model? It would suit our commercial and industrial cluster micro grids.

  4. Well Peter, you are a researcher. Prove it.

    If what you postulate is true, why aren’t there thousands of diesels generating power around the USA? Diesels are a fraction of the total energy generation resources in mainland USA. For several very good reasons.

  5. I have not read the Stanford study, but the discussion above appears to be limited to the prospect of supplying 100% of the electrical needs of the model island. It would be good to clarify if end uses requiring heating/cooling or mechanical energy are included… and consideration given to using non-electric renewables for them. Of those end uses needing electrical energy, were end use energy efficiency options considered first? The relevant question is… How much of what remains then (after addressing efficient end use) as the electrical need can be supplied by renewables (or hybrids) in a cost effective manner? In such an approach, the end use (demand side) is given primacy over the supply sided question.

  6. I also believe that past the 77% mark on the above scale is very costly, but to combine the uses of many hybrids along with diesel in moderation is the beginning, we must begin with smaller systems that can perform independently and be linked to the larger systems that can flow both ways to meet the demands as the demand moves from one area to another, WE as a global society must get past the massive dependence on burning anything that cannot be harnessed transformed and used to generate energy, WE must start building these systems to work together and as independent power gen systems, using diesel as the standby, we will always (in my lifetime)have the dependency on diesel usage but we can start the process now of converting to combining hybrid systems to become linked together as time permits by locations and systems sizes to the 77% – at least till Luke teaches me how to use my force.

  7. This is an interesting discussion. Don’t forget that once upon a time the big concern was 30% being the big hurdle, so its nice to see the 100% being discussed. Particularly relevant for small island nations.

    I would just like to point out however that most arguments against the viability are assuming models of energy use and demand of yesterday. Getting to 100% requires a whole new thinking in this area and we must consider extensive demand side management. If you consider that a system can survive without power for just 1% of the time, it makes a huge difference to your model. A power cut on a remote Island or anywhere that isn’t a sprawling metropolis is not the end of the world, and is often common place..a part of life. We must learn to use energy in harmony with resources….ask anyone that lives off grid. 100% renewable needs new models and new ways of thinking, but if we start approaching it I have every faith that it is easily achievable. LCOE by the way is a fool of today’s economic thinking that will go only as far as you can throw it and every situation is different.

  8. Again I ask, Why Diesel Fuel? This is an oil byproduct. Diesel fuel cannot be combusted as efficiently as LPG or CNG.
    Ten days ago it is being reported that 400 ppm CO2 in the atmosphere has been reached.

    What fuel source combusted out of necessity is better for our environment?
    What fuel source can be consumed more efficiently?

    Why diesel fuel?

    1. Diesel fuel can be made from vegetable oils. When the first diesel engine was patented it was designed to run on vegetable oil. Germany’s Parlament Building in Berlin is partially powered with diesel generators powered with canola oil which they call rape seed oil. Feedlot, sewage treatment, land fill, coal seam bed and permafrost escape of methane should all be caputured and either burned or used for making plastics. Alcohol and methane can be made from biomass. Hydrogen made from surplus electricity from wind, solar, tidal, geothermal, and hydroelectric is probably the cleanest of fuels that can be burned or used in a a fuel cell. The less CO2 and methane release the better. We have to tax carbon and sequester any CO2 released. But biodiesel, bio-alcohol and bio methane are essentially renewable fuels. They need to be a part of the mix but should be avoided until all other sources of fossil fuel and nuclear energy are stopped from use.

    2. “Again I ask, Why Diesel Fuel?

      With island and remote microgrids, the difference in delivery cost between liquid and gas fuels (compressed or liquified) becomes a major factor.

      1. Alcohol can be made from lots of different plant material, including dead floral beau-ques. Methane or natural gas can be used as a fuel after being scrubbed of water, benzene, toluene, xylene, butane and ethyl ether. propane, ethane and other VOC’s (Volatile Organic Compounds) and gases. Methane can be made in a digest-er or extracted from animal feedlots, abandoned coal seam beds and sanitary landfills and compost piles. Hydrogen can be made by separating water or from methane. Special fuel cells can take methane combine it with oxygen from the air and make electricity, water and carbon dioxide, without burning it. So we have many sources of liquid and gaseous fuels that could be portable and renewable.

      2. Hi John,
        diesel is simply what’s out there. Go to any remote Island around the world and chances are they will have two or three big old diesel engines. These things tend to last and stay with us, so we model them…easy to say why diesel, but this is what is out there. They are robust, easy to fix and very common and can be adapted. Sure they would probably switch if they could afford it, but often it is difficult to find even one supplier, let alone one of liquid gas (depending where you are). The cost of replacing plant in a remote place is double that or more, than say near a major center. Two thirds of the cost of goods in the Pacific Islands goes on transport just to get stuff out there. We see a big drive in the Pacific now towards using coconut oil, so again a replacement for diesel. I acknowledge the difference in delivery for liquified gas, but finding willing suppliers and changing out kit is not a typical solution for most. Investment now is going towards renewables and this is the major driver of changing plant, replacing fossil fuel, as opposed to switching the type.

  9. Diesel generators are more ‘flexible’ than a simple cycle gas turbine installation? How do you define ‘flexible’?

    When you are not on an island burning imported diesel fuel, they are only selected in cases where they are only expected to run in emergency situations.

    The installed cost of diesels are marginally lower than the installed costs of a GT, but their O+M costs are very high. Airplane based GT designs have virtually zero O+M costs. I’ve never heard of a case were diesels were selected over a GT because the diesels have quicker ramp rates or start faster. I’ve only heard of cases where diesels were selected over GTs when they were expected to run less than 5% of the year.

    Island installations are a very small market, with unique needs. The needs on the mainland are very different, and result in very different results.

  10. What you have graphed is called the 80-20 rule of thumb, ie, 80% of the project can be done with 20% of the cost, and the remaining 20% with 4 times that. However, your assumptions are “dated”, and/ or not necessarily the correct assumptions. Energy is economics, and you are taking a sterilized, non specific case, which has only limited (desert island type of context). It’s the interaction usually synergistic or unintended consequences that screw up isolated assumptions model. Example, if RE gave your model the ability to make aluminum ingots and thus stock cheaply (12,500 kWh/ MT to make aluminum), versus buy it at market price, reflecting a multiple of Iceland’s RE cost (Geothermal $.04/kWh or less), you could not manufacture competitively. If the nation’s GDP reflected aluminum stock in manufactured products, and was knocked out that industry without a RE resource, I don’t see how your model would that realistic set of circumstances. .

  11. Why diesel fuel? Why not LPG if the fuel has to be transported in?
    Natural gas and LP Gas are fuels that can be consumed so efficiently that the heat energy can even be recovered from the exhaust and utilized. The Water will be created when the heat energy has been recovered, and this distilled water is also very usable.
    We want near 100% energy efficiency, that is why we have to realize the benefits of using these “clean” energy sources.

  12. Predicting costs accurately matters. We can no longer talk about producing energy without talking about all the costs involved. Nor can we continue to compare energy investments without tallying their true costs. When we do we will make different choices. What do your levelized costs include? Any of the below?

    External’ costs have inflicted substantial damage. Toxins released from coal power plants account for 80% of the mercury found in the fish we can no longer eat every day. Released particulates keep Richmond in the running for the city with the worse asthma rate in the country, and, in the aftermath of hurricane Sandy, New York State recently warned that it cannot predict the need for disaster funds accurately as climate change raises the sea level and flooding occurs in unheard of places like the subway system.

    Health and environmental damages are financial costs too; costs paid by the state with our tax monies, or by the increased costs of our damaged health. Because these costs have been excluded from utility calculations, they have effected the decisions of our utility and ultimately, its regulator, the State Corporation Commission.

    Add to the above the 100 years of tax expenditures the fossil industries have used ….

    1. Fair enough, but we are looking at this from the perspective of real projects in real time. We are looking at the direct costs involved, and saying that in many cases, particularly on remote grids, that 80% is extraordinarily compelling. You don’t need to bend over backwards or try to levelize indirect costs, which are not relevant to a particular project at a particular time – 80% works. 100%, however, adds huge direct costs that we think would be better spent creating more 80% situations. Please don’t misunderstand us – we spend our lives in pursuit of maximizing renewable energy! We are huge proponents of it. And we are trying to make a very practical economic point about direct, measurable project costs. If we start there, it gives us much breathing room to create the much more difficult issues of policy and regulation that can take years to realize.

  13. The answer to the quandry of renewable energy storage is just now becoming clear with people starting to understand that electric vehicles are not just a way to save money and the environment. By changing the way we look at electric cars from somenting we charge up at night and drive to work and back into a mobile renewable energy storage device that is charged during the day with excess RE generation, The batery pack is taken “home” where it powers the house and possibly gives the grid extra capacity. Smart meters and smart phone technology make all this possible….

    There is no better way to excelerate EV utilization than to subidize their use through “smart” battery storage system technology.

  14. I think we need to look at this in a wider and more comprehensive way. Germany fuels some of it’s diesel generations with canola oil or rape seed oil as they call it there. When the diesel engine was invented, it was designed to run on vegetable oil, not oil derived from petroleum. What about hydrogen fuel cells? What about battery storage? What about conservation and more efficient buildings, machines, appliances, heating, cooling and mass transit? What about pumped storage hydroelectirc? What about tidal, wave, river current, geothermal, and wind? What about biomass and co-generation? Each state, region and nation on Earth will have a different mix of sources of power. We need to change the laws so people can put PV’s panels on their roofs and have smart metters to sell it back to the electric grid. This is a complex problem but it can be solved. What is stopping it is not the science or the technology but the greed of the centralized electric utilities, fossil fuel and nuclear power industries whose profits will suffer. They are buying off the politicians with their unlimited campaign donations, which control the regulatory agencies. They also brainwash the public that we need fossil fuel and nuclear fuel for a “tranistion” or for our “energy independence.” They also say that we don’t have the technology or infrastructure to transistion to renewable energy. They also use fear tactics telling the public that if we switch to renewable energy and stop using fossil and nuclear energy, that we will all “starve and freeze in the dark.” Does anyone remember the 50’s and the 60’s when the electric and nuclear industry said that nuclear power would be “too cheap to meter?” Despite their money and power, people are switching to renewable energy on their own. With or without tax subsidies and regulations, I hope people can continue to transition to renewable energy. What if they had a power grid and nobody plugged in?

  15. While your model may work for displacing diesel engines on an island, I have serious doubts, with it’s current assumptions, applicable to the situation in the US.

    First, I’ll admit I have not read the Stanford study. I also believe that the challenge is not technical. We certaintly have wind, solar and energy storage technologies that have been deployed and could be deployed to accomplish a 100% renewable energy goal.

    The issue is primarily an economic, and political. There will have to have the political will to cover a significant portion of an area with solar farms and wind turbines. Biomass could possibly be expanded to make a larger contribution. I’m not sure how the authors classified hydropower, but hydropower being ‘renewable’ is a controversial topic.

    The early retirement of power plants, the necessary grid upgrades, the investment in the new renewable sources of power and the necessary energy storage to make renewable energy reliable would have a significant impact on electric prices.

    Significantly higher prices than in other areas of the country would likely lead to many industries leaving. Higher prices would also make energy conservation more attractive, for those who could afford it. Although the poor would be likely to suffer disproportionately from significant increase in prices.

    I would offer a more economic approach would be to meet all increases in demand and the orderly retirement of generating facilities at the end of their useful life would make more sense. Yes, this would take longer, but the total cost impact of switching to 100% renewable energy in the short term is likely to me enormous.

    1. The cost of going to 100% renewable is even greater on grids that do not use diesel generators. Diesel generators are much more flexible than any other conventional generator. We all agree that it is not a technical problem. Our point was that the last 20% (maybe it would be 30-40% on non-diesel grids) requires so much storage that it is several times more expensive. We would accomplish more by getting everyone to 60-80% than trying to get a few systems to 100%. It is true that we are limiting our discussion to solar and wind. A renewable liquid fuel would allow a system to go to 100%, but that fuel is probably more valuable as a transportation fuel.

      1. It is terribly misleading to use island power as a proxy for all power systems. A simple island power system cannot take advantage of the benefits of aggregation and load shifting that are available for a large grid system particularly in North America. Add load shifting and EV storage (first as a sink and as batteries come down in cost as a source in V2G) and 100% grid is looking promising. Note also that externalities are not fully monetized in the current system. If that occurs through carbon tax or cap and trade the economics tilt even more in the favor of zero carbon electricity and transportation and some HVAC as heat pumps become more widely used.

        1. You raise several interesting points, thanks for taking the time to comment.
          We think it is easier for an island or other microgrid to get to very high penetrations than it will be for a continent-scale grid. In fact, there are several islands that are already claiming to be 100% powered by solar, or wind, although they contain a caveat that fossil generation is still used as a backup, so they are really 90+%. Maybe it is silly to quibble about the last couple of percentage points. Tokelau, King Island off the coast of Tasmania, Australia, Kodiak and St. Paul islands in Alaska among many others all operate 100% renewable for extended periods. But not for the whole year. One of the things that these islands have demonstrated is that there is a fundamental difference in the controls and power electronics that is required when operating diesel-off. It is really a quantum leap increase in sophistication. For a large utility system, the increase in control complexity is even greater. Very small systems (think of a cabin on a mountainside or a solar home system in a developing country) are usually 100% renewable, but with a single user, they can manually control their loads in a way that is not possible for a utility company. Perhaps the most important point is how much easier it is to manage loads in a microgrid than for a utility company with millions of customers. You are absolutely right that load management, EVs, expected improvements in storage and carbon pricing all help make high penetrations more economical. We do a lot of modeling of different sized systems. That analysis shows that almost all systems can cost-effectively add substantially more renewables, thereby reducing both the cost of energy and carbon emissions. In addition to the factors you mention, biomass, hydro and geothermal, where an appropriate resource is available, push the optimal level of renewables even higher.

          1. Dear Lilienthal Sir,
            I think its time to update this post. The advances in storage and other renewable technologies seem to disagree with the post….and partly thanks to your innovative HomerPro.

Comments are closed.