Hard Solar Realities in Puerto Rico
Hurricane Maria annihilated Puerto Rico’s power grid. The power generation systems that were already in varying states of disrepair before the storm are now critically wounded and the transmission lines that take that power to the homes of 3.5 million Puerto Ricans are 75% destroyed. Typically after these disasters, power companies simply rebuild them — fix the power plants, lines and transformers and turn the switches back on. That is not going well so far in Puerto Rico.
Enter the Green Gods: Elon Musk famously spoke with the Governor of Puerto Rico this week to discuss replacing the aging systems with shiny new solar panels and Tesla PowerPack battery systems. Environmental pundits began suggesting the same. Replace the dirty, expensive fossil fuel plants with solar. The island has suddenly become a perfect setting for these dreams — small, solar microgrids to run hospitals and feed air conditioners (and yes, to power lights, TV’s and Sony Playstations.) Solar is the answer, the time is now. What an ideal place to prove the viability of modern renewables and the newest, shiniest battery systems (and for Tesla to do a little marketing in the process.) Let’s imagine what it would take to make this dream come true. What would it really take to mostly power Puerto Rico from solar energy?
Puerto Rico is an island nation of over 3.5 million people, and its power company, PREPA, has 1.5 million customers (meaning, actual meters on actual buildings.) The total capacity of all the major power generation systems in Puerto Rico equals 5.359 billion watts, or Gw (GigaWatts.) Like most smaller island nations, it has no coal mines or gas wells, so the fuel for power mostly comes in on ships in the form of fuel oil, diesel, coal and liquified natural gas. (Puerto Rico does have about 200 million watts (MegaWatts or Mw) of solar and wind generation, less than 5% of its total generation — but these sources produced less than 3% of the power used in 2016.) Overall, the island used about 20 billion KiloWatt hours of electricity in 2016. Said in another way, that’s 20,000 GigaWatt hours (GW), a common measurement of power generation for big power plants.
Puerto Rico has very little hydro power, (although the potential for more is certainly there) so it employs a mix of small and large power plants burning coal, oil, natural gas and diesel. These generators are on average 45 years old, falling apart, unreliable, dirty, and expensive to feed with shipped-in fuel. What it would really take to fully replace these old generators with new solar power?
First, consider the resource — the sun. Perhaps you’ve heard the statistic that enough energy from the sun hits the earth each day to power civilization for 1,000 years. It’s true. The problem is that star power is fickle. The sun has a habit of going down at night, and clouds — especially on an island surrounded by warm water — often cover the sky. The percentage of time that a solar or wind system produces useful power averages 26% for wind and 13% for solar worldwide (This is called Capacity Factor).
But the sun shines 12 hours a day in Puerto Rico doesn’t it? Mostly, yes, but solar panels only output near full power when the sun is nearly directly overhead — so not in the morning or evening — and not when it’s even a little cloudy. In other developed countries, where wind and solar generation exists, they are used as a secondary source — when the sun goes down, all large power grids rely on fossil fuels, nuclear and/or hydro (dams) to keep the lights on, the heart surgery lit and the AC running.
Without lots of fossil fuels, hydro or nuclear for backup, what do you do when the sun isn’t shining? Enter the batteries. We have advanced battery technology massively in the last decades, which has just recently, and barely, made electric cars viable. We still — Tesla included — just suck at building batteries to power big things like cars, much less cities. However, recent battery advances, mostly in terms of cost, not technology, have enabled a few small working examples of grid scale battery storage, so let’s imagine scaling those up for Puerto Rico’s needs.
Tesla recently built battery backups systems (albeit at a small scale) using its “PowerPack” systems for grids in Australia, Kauai and Southern California, each in the 80–100 MegaWatt range, using the lithium batteries developed for its cars (100 MegaWatts sounds like a LOT right? Back to that in a minute.) So, solar and wind pumps power into the batteries when they are running, and the batteries feed power back to the grid when they are not, pretty simple in theory. Efficiency on the PowerPack battery systems is pretty high — Tesla reports 92%. One big problem with lithium batteries, however, is a lack of longevity. Does the lithium battery charge in your phone still last as long now as when it was new? Mine doesn’t. Like, at all. But in the name of optimism, let’s assume the PowerPack’s efficiency and longevity are sufficient for Puerto Rico’s needs.
How much would it cost to put solar systems and battery backups in place to power most of Puerto Rico, assuming perfect efficiency? Let’s start with the solar panels themselves. We’re going to use PV panels, not solar thermal, as they are proven, available, and, easy to estimate pricing.
We can look around the world for examples of very large PV solar plants in areas with a DNI — a measure of how much solar energy actually hits the ground in a given location — that are similar to those in Puerto Rico. Puerto Rico’s strongest DNI values are on the north and south coasts, at about 5.7. This is similar to the newer Yanchi Solar Park in China, a 6,671-acre array with new, highly efficient technology. Yanchi produced 525 Mw hours in 2016 from 380 MegaWatts of PV solar panels.
We need 20,000 Mw hours for Puerto Rico, so we divide 20,000 (the total need) / 380 (the total Yanchi plant output) = 52.6 . Meaning we need 53 of these massive solar farms, which would total 140,000 acres of panels, to get the power we need to keep the lights on in Puerto Rico year-round. The total acreage of Puerto Rico is 2.24 million acres. Realistically, solar plants there would be distributed — lots of little plants, rather than a few giant ones. Even so, sourcing 140,000 acres — or 218 square miles — of open land on that island would likely require clearing virgin rainforest. And require a significant investment to purchase the land in the first place.
Solar PV panels are getting cheap as you may have heard . As of today, they cost about $1.50 a watt for these big utility plants. So, to install the 20,000 Megawatts we need , that cost of the panel systems (not including the land) is about $30 billion dollars.
So far, we’ve found 140,000 free acres of appropriate land and shelled out a cool $30 billion for the panels. Now back to the batteries.
In 2016, Puerto Rico’s annual energy consumption translated to about 54 million Kw hours per day. That means we need to be able to store and deliver that same 54 Million Kwh daily in rain or shine, morning or night. How long do we need to be able to rely solely on batteries? It rains fairly often in Puerto Rico — and solar panels don’t produce much useful power on cloudy days. Let’s size our system to keep the island running for up to three days without sun. After that, the power goes out until the sun shines again. That means our battery system needs to contain 54 x 3 = ~150 Million KwH of storage.
The example of the Tesla Australian PowerPack grid battery system costs $250 per kilowatt hour — Elon Musk himself announced the cost on Twitter recently. Using similar batteries to run all of Puerto Rico for three days would cost over $37 billion.
The global manufacturing capacity for these batteries, even with Tesla’s new GiGaFactory , does not appear to be able to supply this demand right now — but let’s ignore that for the sake of this thought experiment.
For as little as roughly 67 billion dollars, we can power Puerto Rico with solar, including a battery back-ups for up to three days. The panels will last perhaps 25 years, and the batteries, well, nobody knows exactly, not even Tesla. Tesla’s public warranty is 10 years.
What’s $67 billion when we’re talking about clean energy?
Well, the GDP of the entire island is $100 billion per year. And, Puerto Rico is already about $70 billion in debt. 45% of the population lives below the poverty line. If you are the Governor of Puerto Rico — can and should you get a $67 billion loan for this new renewable power system? The island’s credit is pretty sucky, so it likely can’t get a loan. Perhaps the U.S. should loan the money, or just give it? Is $67 billion dollars for a pure solar play the smartest way to generate energy for Puerto Rico? Is that mega investment the one that would most improve quality of life for Puerto Rican citizens?
So what should Puerto Rico do to bring full power back in the wake of hurricane devastation?
The most sensible strategy now is a repair of the existing grid and generation, along with the modern upgrade to their fossil fuel facilities, and/or perhaps more privately-built, for-profit power plants, like the two modern combined-cycle gas plants the island has now. This is the best Puerto Rico can hope for without significant external investment. That may also include some smaller, high-profile Tesla battery/solar systems to help backup critical infrastructure like hospitals, which will be celebrated as a major move towards renewables, but will make little difference in their burning of fossil fuels.
Hydro is perhaps the most attractive renewables option for Puerto Rico, reducing reliance on dirty imported fuels, and generating power largely without carbon emissions. Hydro power would not require batteries — the water behind the dams is the battery. But it would have all the downsides of hydro as well : ecosystem and habitat loss, popular opposition and reliance on fickle rainfall and waterway supplies. I have done little research on hydro power in Puerto Rico other than finding data that show there is little of it, and that the resources do exist to build large dams to produce power.
Wind generation generally costs materially less than solar, but the needed battery systems would remain largely as described above if not backed up fully by fossil fuels, nuclear or hydro. Current installed costs per Mw for wind (on-shore) are about the same as solar PV per watt — about $1.60 — however, they produce more usable power for the same watt, as the wind tends to blow more per day than the sun shines, all other things being equal. As such, we can assume that a total wind, or hybrid wind/solar solution in Puerto Rico would cost a bit less than solar, and the battery backups requirement a bit lower, due to the mix of the two sources — but still in the $50 billion dollar range.
For about $7 billion (a pessimistic number), we could build a latest generation nuclear power plant that takes up 200 acres (vs. 140,000) and will power the entire island day and night, rain or shine, through Category 5 hurricanes, for 50–100 years. Of course, nuclear has its own set of downsides — radioactive waste, and safety concerns — yet a strong argument exists that those problems are preferable to the impacts from other sources of generation. Environmental impacts may be less than large dams, and are almost certainly less overall than burning fossil fuels. Given the high prices the people of Puerto Rico pay for electricity (higher than any U.S. state except Hawaii) a private investor may be willing to build a plant with loan guarantees and/or insurance from the U.S. government.
The solar story is a much more appealing fairy tale.