Hydrogen in the Hills?

Every few years, the zeitgeist shifts and a new set of pundits begin promoting the future prosperity offered by the dawning of a new hydrogen based economy. I recall reading about the exciting near-term development of commercial photocatalytic hydrogen production in my AP Chemistry class about 40 years ago. When I worked at a bookstore in Boulder in the early 1990s, I recall a number of books exclaiming the future hydrogen economy with hydrogen pipelines replacing natural gas pipelines supplying clean burning hydrogen gas to every house. Walt Pyle wrote a few articles for Home Power Magazine (free registration required) about hydrogen, including “Cookin’ On Hydrogen Stove Burner Conversion” in HP #33 (Feb/Mar 1993) sandwiched between an article about PV mounting and an introduction to transistors. The first link above, “The Solar Hydrogen Chronicles” is worth a read. More recently, there has been quite a bit of ink spilled about Green and Blue (and various other shades) of hydrogen based on the source of the hydrogen, with so-called green hydrogen produced with electrolysis of water powered by solar photovoltaics, and blue hydrogen from cracking hydrocarbons (such as methane) liberating hydrogen atoms from their carbon molecular prison.

Then, last month I saw a heated exchange on social media insisting that EVs were dead and the age of hydrogen cars was upon us. Nevermind that hydrogen cars have been available (albeit in limited markets) even predating Tesla. Then after extolling the virtues of Hydrogen cars (primarily that they take less time to fill, and that the filling station model would be preserved) it became clear that the advocates either did not realize that the current crop of hydrogen cars were just EVs with a 10000 psi H2 tank and a fuel cell instead of a battery. Other than nostalgia around visiting the gas station “as we always have” I am not sure I see a big difference. One smug respondent seem to think that not being able to charge in one’s own garage was a feature (he obviously was not aware that Honda developed such a unit in 2007 that uses natural gas, i.e. blue hydrogen, or Honda’s Solar Power station i.e. green hydrogen). Others seemed to be focused on the fact that hydrogen *can* be used to power an internal combustion engine (but the ICE is significantly less efficient than the hybrid EV approach.

About a year ago, I posted about Vanadium Flow Batteries as a possible long duration energy storage technology. Sadly, this technology continues to remain unavailable in the residential off-grid market. And while a few other flow batteries based on different chemistries have seen some limited availability, they are more focused on replacing lead acid or lithium battery chemistries, than providing for the long-term seasonal bulk storage that becomes necessary in far northern (and far southern) regions.

In my periodic review of updates for this technology, I did discover that an older technology, one that has been around for over 100 years, is starting to see some general availability in the long-term residential market, at least in Germany. As you can probably guess, I am referring hydrogen, specifically a closed-loop hydrogen system. Closed-loop in this context just means that hydrogen is produced by electrolysers that produce H2 gas from water, stored either in compressed form (perhaps also in a metal hydride) or liquified, and then converted back into electricity when needed by a fuel-cell with water as a waste product. The Picea2 system from HPS, provides an integrated solution that can supply 15 kW of power and store 1.5 MWh. Two systems can be combined for a total of 30 kW of power with 3 MW of seasonal storage and 30 kWh of short term LiFePO4 batteries. Complete solutions including PV panels, inverter, batteries, electrolysers, fuel cells, and fuel tanks start at about $100K. But they are only available in Germany. As far as electricity storage, this system is only 40%-60% efficient, but can also produce heat that can be used to heat spaces or for domestic hot water.

So, what problem I am trying to solve that cannot be solved with my existing technology?

The issue is winter energy production and storage. We do have an abundance of power during the summer, at least on clear days. I would like to use that stored energy, rather than running the diesel generator in the winter.

Consider that excluding climate and weather, just focusing on daily solar irradiation, we only get about 4 hours at winter solstice. And that winter sun, only rises to an elevation of 4 degrees over the horizon for solar noon on the winter solstice. In December, even with solar panels mounted at a 86 degree angle (to be “normal” or perpendicular to the sun) there is just not that much radiation to harvest over the few hours the sun is up, just barely, over the horizon. And that is on a clear day, assuming no other ground or sky obstructions. Add in overcast winter days, tall trees to the south (we live in a forest, after all), and snow storms, all in all, it is difficult to produce much power in November, December, January, and the first half of February.

On average, we consume about 33 kWh per day, which is 1 MWh per month. We would need to store 3-4 MWh to cover our average usage over the winter months assuming no solar production at all over the winter. Of course, this is just considering electric usage. We also use heating oil to heat the house and garage, and propane for cooking, some clothes drying and domestic hot water. To completely eliminate fossil fuel use would require going “all-electric” perhaps using waste heat from a hydrogen storage system, or some form of solar thermal collectors along with heat pumps to provide heat and domestic hot water.

Consider that at solar noon on the summer solstice, the solar altitude rises to 51 degrees above the horizon (which means a solar tilt angle of 39 degrees). There is 20 hours of sunlight. The problem in the summer is that PV panels produce the most power when they are perpendicular (or normal) to the rays of the sun. Over the course of the day, the elevation rises from 0 degrees at sunrise, up to 51 degrees at solar noon, and then back down to 0 at sunset. Even worse the azimuth or heading swings from about 30 degrees (that is North North-East) at sunrise, to 180 degrees (due South) at solar noon, and then over to about 330 degrees (North North-West) at sunset. So, my south facing panels are in the shade until about 10 AM and then go back into the shade about 6:30 PM.

You may recall that I mentioned an instrument called a pyrheliometer that measures direct normal irradiance (DNI). This is the solar radiation received per unit area by a surface held perpendicular to the sun. This could be a 2-axis tracker. I also mentioned another instrument called a shaded pyranometer that measures diffuse horizontal irradiance (DHI). That is the solar radiation that has been scattered by the atmosphere and (because it is shaded) does not include the direct component measured by the pyrheliometer. These two values can be used to compute the global horizontal irradiance (GHI).

When the PV panel is not perpendicular to the sun, it can only receive the diffuse radiation (DHI) from the atmosphere. This is usually significantly less than direct radiation. There are several ways to mitigate this issue, including setting up east and west facing arrays to complement the existing south-facing array to take advantage of DNI during specific times of the day and provide some additional power in the morning and evening, using bifacial panels to capture some of the reflected light, or even using 1 or 2-axis tracking arrays that move the panels to be perpendicular to the sun at all times, thus again benefiting from direct normal radiation. During periods of cloud cover, the direct rays of the sun are obscured, causing the DNI to reach zero, with only DHI available.

NREL in Golden, Colorado has a Kipp and Zonnen RaZON+ that provides both of these instruments and a data logger that shows plots of DHI, DNI, and GHI overlaid on each other for specified days. It is interesting to look at the results of a partly cloudy day, such as March 13, 2024. Here you can clearly see the DNI become obscured and the DHI become the only component of GHI.

Another invaluable resource is NREL’s PVWatts Calculator. Using this tool, which aggregates meteorological and solar irradiance information from many sources across the US, I ran some calculations on how much electricity I could generate from various large arrays. For example, here is the model I produced for a 42 panel SAT-Tracker dual axis array with QCell Q.PEAK DUO XL-G11S/BFG Bifacial 600 Wp panels. Here you can see that even with the biggest, most advanced bifacial panels mounted filling a large dual axis tracker, we cannot generate enough power for at least two months out of the year.

Unlike the Vanadium system I discussed last year, there are hundreds of these residential hydrogen-based systems in use today, sold as commercial systems with engineering and support by HPS in Germany. There even are more systems that use the same technology that have been built as custom engineered projects including the hydrogen house in New Jersey, using products from suppliers such as Enapter, Efoy, and Ballard Power Systems.

The next challenge is H2 storage. While it is possible to generate more power in the summer than I could use or store with batteries, in order for this energy to be useful and to eliminate the need for running the diesel generator in the winter, 3 MWh is the minimum requirement, and 6 MWh (or even 9Mwh) would be preferable in order to also ride out periods of overcast skies and inclement weather during the rest of the year and perhaps begin to convert some of the applications for heating oil and propane and/or provide for a future electric vehicle.

So how much hydrogen will we need?

Each kilogram of hydrogen stores about 33 kWh of electricity, so 100 kg of H2 stores 3.3 MWh, 200 kg 6.6 MWh. A couple of hundred kg does not sound like much, and as far as energy density, is comparable to the energy density of gasoline by weight. But hydrogen is a very light, in fact the least dense element by volume. Storing 200 kg of H2 at atmospheric pressure would require 84660 cubic feet! As a liquid, a more manageable 2825.6 liters (675 gallons). Considering that we store 1500 gallons of diesel and 1000 gallons of liquid propane, 675 gallons seems reasonable, until temperature is considered. Hydrogen cannot be compressed into a liquid, it must be cryogenically cooled to 20 degrees Kelvin, that is VERY cold, -253 degrees C (-423 degrees F). That cooling process may consume 10 kWh of the 33 kWh energy (cf. Leachman 2020 https://hydrogen.wsu.edu/2020/10/06/could-smaller-hydrogen-liquefiers-be-better/). There is now a commercially available light scale hydrogen liquification system from GenH2 that can produce 2-20 kg of hydrogen per day.

Other options include high pressure storage. Tanks and compression equipment to support storage up to 10000 psi (700 bar) are readily available. At this pressure, the required volume is about 40 liters per kg. So that is 3500 liters or about gallons. Due to interest in hydrogen as a automotive fuel source, there are a variety of high pressure hydrogen cylinders available.

So, is there a hydrogen future for Noctiluca? Maybe, but probably not for a while. Although this technology is certainly more mature and available than Vanadium fuel cells, and is more scalable for long term / seasonal storage than lithium batteries or zinc bromide fuel cells, home sized fuel cells and electrolysers are still comparatively expensive. And as much as I would like to replace the 500 gallon diesel tank sitting behind the garage with a 3000 liter cryogenic tank, I think the market needs a few more years to mature. In the meantime, I can focus on shorter term projects while I continue to track hydrogen and other long duration storage options.