Will Distributed Solar Drive Utilities into Bankruptcy?
Electric utilities today look a lot like newspapers in 2000: Too much debt in an industry primed for disruption.
Speaking at the Economist's Intelligent Infrastructure Conference, Brad Tirpak, Managing Partner at the private investment fund Locke Partners made the case that electric utilities are as woefully unprepared for the coming disruption of cheap, distributed solar power as newspapers were unprepared for the disruption of the Internet in 2000.
He outlined the following parallels:
- Both had long been considered to be sure-fire businesses with dependable income.
- Both took advantage of the seemingly dependable income to load up on debt.
- Both face disruption from a disruptive technology (the Internet, and distributed generation and efficiency) with the potential to undermine their businesses.
Newspapers have not gone away, but as readers and advertising increasingly migrated to the Internet, circulation numbers dropped. When a company is loaded with debt, a small drop in revenues is magnified into a proportionately larger drop in profits. To stay solvent, newspapers had to raise prices.
Rising prices drove more readers away, starting the cycle all over again, and eventually leading to bankruptcy for many of the papers. As you can see from the chart below, many of those papers that survived without bankruptcy lost most of their stock market capitalization as more and more of their income was needed to service their debt.
The Price of PV
Mr. Tirpak expects a similar story to play out in utilities. As solar becomes cheaper and reaches grid parity, installations will grow rapidly.
Edward Fenster, CEO of SunRun made the case that we don't even need further decreases in solar photovoltaic (PV) panel prices to reach grid parity solar even without the federal subsidies. According to Fenster, solar panels currently cost $1.65 per watt, but total installed cost is about $5.50 per watt. While some of the extra cost is Balance of System (wiring, inverter, mounting), the majority is labor and permitting. In Germany and Japan, permitting and installation are only $1.50 per watt: Fenster believes we can get there too by doing away with local permitting on standard installations ($0.50 per watt reduction) and using greater scale and operating leverage ($1.50 per watt reduction.)
Those reductions would lead to an installed cost of $3.50 per watt. According to my calculations, that would lead to a 30-year internal rate of return of 4% (IRR) given a 20% capacity factor and a $0.13 cost of electricity per kWh. If we assume any electricity price inflation at all, the IRR increases with it, and a $3.50 per watt PV installation looks attractive at any interest rate below the IRR. We can also safely assume that there will be further reductions in both panel prices and in other system component prices.
What Might Happen to Utilities
PV will probably reach grid parity in the next few years, through a combination of rising utility prices, increasing returns to scale in installation, and cheaper balance of system costs. If this then leads to rapidly growing PV installations, will it undermine utility revenues, as the internet undermined revenues at newspapers?
I think the analogy is based on a misunderstanding of both the scalability of distributed PV and the utility regulatory environment.
First consider the regulatory environment. Utility regulators are charged both with ensuring that utility customers get service at a reasonable cost, and also that utility investors will continue to be willing to provide capital for necessary utility investments. If the rapid spread of PV were to threaten utility solvency, regulators would take action to help the utility maintain solvency.
Mr. Tirpak understood this, but made the assumption that the only action regulators could take to protect utility solvency would be to raise prices, which he assumed to mean the price per kWh of net energy used. If this were correct, then we would indeed see the vicious cycle of increasing rates and declining volumes that has undermined the solvency of newspapers over the last decade.
It's not all about cents per kWh
Regulators have other options. First, they can allow the utility to cut any PV subsidies intended to help the utility reach solar energy targets. If a utility were threatened by too much solar power, such subsidies would clearly be unnecessary to achieve the statutory PV penetration. Subsidies are frequently cut in response to unexpected growth in PV installations. In fact, declining subsidies in response to installation growth are often designed directly into these programs.
Once subsidies are gone, the next step to protect utility solvency in response to PV installation would be to change the structure of electric rates. Although we often think of energy (kWh) as the only thing we buy from utilities, in truth we buy another valuable service: electricity on demand. Even a home with enough PV to produce all the electricity it needs on an annual basis cannot disconnect from the grid: The power must be kept on at night and on cloudy days, and excess electricity needs to go somewhere when the sun is bright.
Electricity storage could be used to take a home entirely off the grid, but such storage would be prohibitively expensive. If a home's average usage and generation is 24 kWh/day (requiring a 5 kW PV installation), then enough battery storage would be needed to get the house through a few cloudy days when generation is greatly reduced. Deep cycle lead-acid batteries typically cost $$200 per kWh, so three days worth of storage would optimistically cost $14,400, or $2.88 per kW of installed PV, making even $3.50/W PV uneconomic.
Since PV does not enable users to do without utility service, regulators can increase the fixed cost of utility service without increasing the variable (per kWh) cost. This price rise will improve utility profits without improving the economics of PV. Other options would be to switch to time of use pricing for electricity, with low prices being charged when there is excess electricity (which would be when PV is operating, since we are assuming a PV glut) and higher prices when there is not enough (dusk on hot summer days.)
In a private email, Tirpak responded to this argument by saying he could not "quantify the support for solar. People hate utilities and love solar. Republicans and Democrats support it. At the end, the [utility regulators] will listen to the public as well as reliability." I certainly have met too many Republicans who hate solar. As for utility regulators (and I've testified before electricity regulators several times), I simply can't imagine them intentionally adopting policies that would drive a utility into bankruptcy.
I can't quantify the public support for solar, either, but I can put an upper bound on it. Residential solar leasing companies like SunRun now can provide solar electricity to customers in seven states for less than the cost of grid electricity, without any upfront cost. They're doing good business, and driving rapid market growth, but most homes in those states still don't have solar: SunRun uses innovative strategies like partnering with One Block Off the Grid (1BOG) to assure sufficient volume. If everyone truly loved solar, they could just hire a call center in India to answer the deluge of telephone calls spend most of their efforts installing panels.
There are natural limits on how much PV can be installed by customers. Many people's homes are shaded by trees or other buildings. Other customers are renters, and so do not have the option of installing PV. Industrial and commercial rooftops are seldom big enough to produce enough power to meet relatively high industrial and commercial electricity usage.
Utility scale installations could produce enough electricity, but such installations need to sell their power directly to the utilities, at much lower wholesale rates. It will be quite some time before solar PV is able to compete at wholesale rates in the absence of subsidies.
Tirpak also lists other potential disruptors of the utility model: energy efficiency, smart grid, LEDs, ground source heat pumps, and cheaper hydrogen. He did not go into detail on why he expects any of these to be significant, but my take is that only cheap hydrogen has the potential to change the story I outline above.
Smart grid, by its nature, is being implemented by utilities at regulators' request: the smart grid will not allow us to do without the grid, since it is the grid. Perhaps Tirpak instead meant microgrids, which are enabled by smart grid technology. While microgrids have the technical capability of cutting the cord to the larger utility, they seldom have the legal authority. A microgrid supplying power to a small group unconnected to the utility would legally be a utility itself, and subject to utility regulators. For the reasons outlined above, those regulators would not allow the formation of microgrids to undermine the solvency of the utility.
The potential for LEDs to further reduce energy use is fairly small. In 2008, I made a weirdly similar (and similarly overblown) argument that utilities might be undermined by the phase-out of the incandescent light bulb. My argument was not that this would reduce electricity sales (which it will), but that it will undermine utility energy-efficiency programs. This will happen because the phase-out of traditional incandescents would make the former stalwart of residential energy efficiency programs, the compact fluorescent light bulb, (CFL) the new baseline. Current LED bulbs use almost as much energy as CFLs of the same brightness, although the technology has the potential to use only 40% as much. But even assuming that LED technology reaches this potential, where a CFL saved 75 watts replacing a 100 watt incandescent, an LED only has the potential (at best) to save another 15 watts: One-fifth of the savings of the CFL when compared to an incandescent. Current technology saves only 2-5 watts over the CFL, at a cost of $40. If the now mature technology of CFLs did not disrupt utilities, LEDs don't have a chance.
Ground-source (aka geothermal) heat pumps (GHP) are already a mature technology, and so are unlikely to see rapidly falling prices like solar. That said, they are already an enormously efficient way to heat and cool a building, and their widespread adoption would do much to reduce energy use. That is why I like GHP stocks. However, GHPs are more likely to be a boon to electric utilities than a burden. GHPs replace heating by natural gas or fuel oil with electricity, adding to utility sales. Just as important, the timing of electricity used by GHPs has the effect of improving utility grid utilization. When heating, GHPs run mostly in the winter and at night, which is just when utilities often have low demand and high generation from wind. When used for cooling, they reduce summer peak loads by displacing less efficient air conditioners.
More broadly, energy efficiency technologies (which include LEDs and GHPs) are unlikely to undermine utility revenues because of the significant barriers to adoption. After all, energy efficiency is already much cheaper than grid based electricity, costing only a few cents per kWh saved. With grid electricity costing five times as much as efficiency already, it seems unlikely that a price shift that makes it cost even ten times as much will make a radical difference in the rate of adoption of efficiency technology.
Of the technologies Tirpak listed, only cheaper hydrogen has a chance of disrupting the electric utility model the way the Internet disrupted newspapers. Hydrogen might disrupt utilities by providing a cheap way to store electricity, which in turn would allow individuals to go off the grid. Yet while hydrogen has the theoretical potential to provide relatively inexpensive energy storage, cheap and efficient electricity storage with hydrogen has not yet even been demonstrated in the lab, at least to my knowledge. That puts any such technology at least a couple decades away from commercialization. I'm not holding my breath.
Given that utility customers are captive in a way that newspaper customers never were, it seems unlikely to me that utility stocks in the coming decade will follow the performance of newspaper stocks in the last decade. Lower prices for and increasing penetration of PV will change the way we pay for utility service, but not free us from utilities all together. Only the advent of extremely cheap electricity storage would allow us to truly cut the umbilical power line, and until we can cut that line, regulators will find a way to charge us enough to keep utilities solvent.
While regulated utilities should weather the coming solar storm, independent power producers (IPPs) which sell their power into the spot market, or whose power purchase agreements (PPAs) expire at the wrong time, might be threatened. This is especially true for IPPs with inflexible generation that cannot easily ramp up and down to compensate for fluctuating electricity supply from renewable sources.
If you're convinced that PV is on the cusp of grid parity and rapidly expanding deployment, don't short regulated utilities, as Mr. Tirpak suggested. Instead, look at IPPs with mostly coal-based generation fleets and PPAs expiring in five years or so.
DISCLAIMER: The information and trades provided here are for informational purposes only and are not a solicitation to buy or sell any of these securities. Investing involves substantial risk and you should evaluate your own risk levels before you make any investment. Past results are not an indication of future performance. Please take the time to read the full disclaimer here.