Monday, September 16, 2019

Renewable hydrogen economy economics


While modern battery systems are becoming the preferred energy storage solution to pair with intermittent renewable (cheapest of all power) energy sources, hydrogen (paired with small battery sizes) is and will be a cheaper and more reliable alternative that offers a carbon free path for all (100%) of humanity/vehicle/industry's energy needs.

Financial discount rate for home energy savings
A discount rate is an interest rate used to standardize financial payments/benefits to a "present value".  As a starting point a riskless rate, usually equal to a government bond in the same currency as financial flows, is taken, and then adjusted based on the project's risk/uncertainty profile.

The value of a grid tied home (solar/storage) energy project is relative to the continued subservience to an electric utility monopoly.  If the utility eventually goes 100% renewables with mega projects, there will continue to be savings from self-generated solar, as long as surpluses can be sold at reasonable rates.  There is high risk/balance of probabilities that utility will be slow to drive down costs through renewable generation, may demand compensation for fighting climate change, and continue to pass down costs related to transmission, delivery, billing, past-nuclear-boondoggles, and profit.

The fact that utility rates will likely increase at the rate of inflation (reasonably expected to be a steady 2%/year), that home energy savings are tax free (while  bond income is not), that energy price and availability certainty (less risk) is obtained, and that the need for home energy is more lasting than government financial sustainability, there is a case for self produced solar to be funded below the standard riskless rate, over 2% (inflation rate) below government rates when great confidence in government sustainability exists.

Most solar projects have payback periods under 10 years (even if 30-50 years of useful energy is produced), and because they pay benefits consistently each year, the comparable bond (which pays back most of its loan at the end) term is half the payback period.  5 year bonds almost everywhere currently have interest rates of 2% or less, and so it is justified to use a 0% discount rate for home solar projects.

Policy:  Financial securitization of completed home solar projects at equivalent to government bond rates would offer homeowners cheaper financing than mortgages, and offer investors secure assets.  Some jurisdictions have negative bond rates specifically to encourage investment. This theme will be devloped in this paper.

For a 50 year solar + hydrogen project, a doubling of amortized energy costs occurs at interest rates of -0.4% (rate possible in Germany ), 2% (Above current US rates, and most other countries), 6.2% (lower than typical risk capital), 13.7% (higher than most sound projects).  This means that if amortized loan payments for a project at 2% interest rate generated electricity costing 2c/kwh, the same project generates costs of 1c/kwh at -0.4%, or 4c/kwh at 6.2% interest rates.  A 0% interest rate would be amortized at 2%, and result in a 1.25c/kwh project cost.

The general case for hydrogen over large batteries
At a steep 60 degree angle, a 10kw solar array in Toronto Canada will generate 30kwh/day in November to January, and average 48kwh/day March to September.  At a steep angle, 10kw solar can take only 25 sq.m (250 sq.ft) of floor area.  30kwh is sufficient to heat and power over 6 floors of this area.  A shallower 30 degree angle will produce 25.5kwh/day in winter, and 58kwh/day for most of other 3 seasons, but 64kwh in July.  Shallow angle solar takes up to double the floor area as 60 degree angled panels.

The right house design for Canada is to meet winter solar needs, and a steep roof is recommended.  In winter, heating/storing hot water and then distributing that water throughout the house is an effective way to use all solar energy.  The rest of the year, however, large surpluses either need to be exported/sold or used up.

The general case for hydrogen is that with a 10kw solar array, 8kw (revised down later) in electrolysis together with 10kwh in battery will use all solar energy and fully charge the battery every single day of the year, while a comparable 50kwh battery does not get fully charged in winter, and only gets useful summer charge, if 50kwh of summer discharges per day every day can find use.

So as long as electrolyzers cost less than 5x more per kw than batteries cost per kwh, the system costs less, provides full use of generated power, and using hydrogen allows complete resilience for any winter event needs.

The value of energy storage is the value of discharged energy.  Whether energy is not discharged, or used to cool home to 16C, because it would be otherwise wasted/unused, it has less value than if there are vehicle or industrial/other buyers for that energy.  The other implication is that long term storage has much less value than regularly dispatched storage, or in the case of hydrogen, exportable/transportable energy.

Home use for electrolysis, oxygen and hydrogen
To produce 1kg of hydrogen (33 kwh of heat/energy value) takes 41-45kwh of electric energy in electrolyzer, and as a byproduct produces 8-13kwh of heat.  The reverse fuel cell reaction takes 33kwh(1kg) of hydrogen and produces 18kwh of electricity with 15kwh heat.  The electrochemical exothermic/endothermic reaction can add about 3kwh of heat to the fuel cell process, and take away the same amount of heat from electrolysis.

The primary benefit of hydrogen electrolysis and fuel cell in home is the heat byproduct.  A fuel cell produces hot "distilled"/pure water.  After heat is extracted from it, the pure water can be stored.  Distilled water is useful for some electrolysis technologies, is best for any piping (heat distribution), and drinking it can relieve kidney stress or over-mineralization in the body.  Any excess heat can be used to improve the efficiency of the electrolysis process, so even in hot summer, you can't have too much hot water.

Oxygen generated by electrolysis can improve air quality, and reduce the amount of air exchange (with outside) in the home which improves indoor heat/cold retention.  Oxygen and hydrogen mixed together and lit with a flame can produce extreme heat.  Cooking is one application, but a simple "over unity" (greater than 100% efficiency) heater can be created by blowing flame over a catalyst.  Oxyhydrogen flame creates steam that is hot enough (with catalyst) to spontaneously disassociate into oxygen and hydrogen, and cycle back into the burn/flame.  A mix of catalyst (platinum) and sand in a metal cylinder can last forever.

Hydrogen generated by electrolysis is an excellent heat transfer fluid.  It is the highest of any gas, and has 1/3 the transmissitivity of water (7-8 times greater than air),  10x the diffusivity and heat capacity of air, the lowest viscosity.  As a gas (with the lowest viscosity of practical thermal fluids), an advantage over water is that it is self pumping to any height and heating/cooling can be gained through compression/expansion.  Low viscosity provides good flow through even thin pipes.  3x the atmospheric density or 3x the flow rate can provide the same heat/cold distribution as water.  A long piping network can double as storage. For very tall buildings, hydrogen distributed for heat/burning (no chimney/exhaust needed) is more practical than central hot water heating because the latter is expensive to pump high.

Financial and technical targets for hydrogen technology
Home electrolyzers must be DC current, priced under $1/watt, and preferably around $0.50/watt.  Of the 2 commercial electrolyzer technologies, PEM's ability to react instantly to variable voltage is not that important with a small battery interface.  Alkaline based electrolysis is generally cheaper.  PEM's ability to use pure/distilled/tap water is a simplicity advantage.  Home heating systems and fuel cells can produce/consume distilled water.  These price and technology targets have been met in large scale commercial electrolysis systems.

Toyota claims its automobile-sized fuel cells (113kw) costs $60/kw.  Residential systems that cost $1000 for 1-2kw would be acceptable.  Automobile-sized systems (Hyundai has 25kw) for multi-family (apartments) and commercial will likely have much better value initially.  Any cooling for fuel cell must be potable water cooled for silence and water heat gain, or through compression/expansion gas cooling.

Electro chemical compressors are very similar to PEM fuel cells, but do not require any catalysts.  These should be produced for under $100/kw, and 95%+ efficient.  300 atm pressure output.  Should double as an expander (generating electricity from pressure)

300atm tanks are about $1/liter in bulk.  Wide distribution network is needed to make this close to effective retail/developer price. Fiber Reinforced Pipe (shipped on a spool) is a useful alternative to tanks or pipelines, and should beat tank delivered costs even when installed on site as a tank substitute. Lasts longer (50 years) as well.

50 year life times for electrolyzers, fuel cells, and compressors should be achievable.

The cost targets are the most important.  To achieve these, it is likely that auto industry R&D in fuel cells will be leveraged, but in the case of a hydrogen producing home, alkaline fuel cells (used in early space flights) that require pure oxygen c/would be suitable.  As is for electrolyzers, Alkaline based technology is currently cheaper.

An appreciated quality shown in some research for electrolyzers and fuel cells, is higher efficiency at less than full current/power.  This is an advantage for solar powered electrolysis and underutilization of electrolyzers.  So, a wide range between peak/"burst" and efficient power rating for electrolyzers is very useful.

Energy control software must be developed to balance use of battery charging, electrolysis, and power consumption.

Hydrogen production vs hot water storage
A solar system without hydrogen electrolysis must put all of its winter surplus energy into heating a large amount of water, then just pump that water throughout the day through ideally under floor piping.   80-95C water can store 70-80kwh of heat per 1000 liters.  At least 2000 liters should  be kept hot to handle poor sun production days, and extreme cold contingencies.  60C water has half the useful hydronic heat as 90C water (for 30c delivery).  6kw of heat pumps in "cascading" mode for high heat lift is needed for the 90C system.  Indoor air to water heat pumps are the most affordable and efficient.  A utility room full of hot water increases the efficiency.  An outdoor/split heatpump/AC can further preheat air and provide additional summer cooling.  Total capacity can be higher than used capacity if cooling is important.

With hydrogen electrolysis, a smaller water system is possible due to, even without a fuel cell, hydrogen as a backup heating fuel that is usable in space/room heaters.  100 liter tank at 300 atm stores 90kwh of hydrogen energy.  Replacing heat pumps with electrolyzers can be (with above targets) cheaper.  Hydrogen tanks use less space than water tanks, and cheaper than a water tank that is 10x larger.

In the "general case for hydrogen section" I suggested 8kw electrolysis paired to a 10kw solar array and 10kwh battery.  The actual recommended system is 6kw electrolysis and 2kw heat pump.  Even if/when the home is unused in summer, a 10kwh battery will allow 1.5 hours of electrolysis before and after the peak 4 sun hours, cycling the battery twice per day for better value from the battery, and with light or heavy energy use in the home, no energy is unused.

Starting in February, even if it is still very cold, there are more sun hours.  Without an electrolysis system, energy should be wasted (for resilience) into overheating water to care against freak cold spells. Resilience exists with hydrogen production. If there is no freak cold spell, there is income potential from sale of produced hydrogen.

A free government policy to significantly reduce cost of renewables and hydrogen production
With a $500 assessment/inspection fee, governments could provide loans to completed/functional renewable energy projects equal to their assessed value (not necessarily cost, though they would be very close, with value tending to be higher than cost in a well executed/designed project), at an interest rate identical to government's low borrowing costs.  While the interest may be the same as any new offsetting government bonds, the project would be cashflow positive to the government, as the loan repayment schedule would follow mortgage (amortization) models with regular principal repayments (faster repayment than the payments government must make on bonds). The government also collects income taxes from the interest it pays, and the profits and wages earned by investment project (= to investment cost). Canada's government accounting system uses the concept of net debt, where assets, especially assets generating revenue/returns, are subtracted from total debt. This program no matter how generous would not increase government net debt. EU member “austerity” accounting should permit these investments without affecting deficits (or their limits)

A completed functioning renewable energy project is risk free investment.  The combination of renewable generation, and storage that includes hydrogen guarantees a pathway for all energy generated to get delivered/used.  Even if long term prices received are not guaranteed (and contracted purchase agreements, may get renegotiated), short term (10 year) prices are forecastable.

During construction, there may be defects and delays and risks.  The industry already has arrangements with 3rd party insurers for long term warranty backing, and in the case of solar, there is weather insurance.  Short term financing by customer or suppliers during development phase has small risk and outlays, but the proposition that customer gets a solar and heating system with no long term cash outflows, and super low financing rates, is extremely attractive, and more attractive than what the current residential solar industry can offer. The US has a 30% tax credit that is only useful to people with high tax bills. A 3% lower interest rate over a 20 year term has equal value to a 30% tax credit fully taken the year it is spent.

There's always some risk of fire and vandalism, but interest rates can be adjusted based on 3rd party insurance against those risks, or an implied interest premium for government provided/implied insurance.  By keeping the loan value smaller than the realizable project value (which can still be higher than cost), you eliminate the risk of insurance fraud motivated arson.

With global 10 year major economy rates all under 1.6% for 10 years, with $15T of investment chasing negative government bond yields.  Bringing the cost of capital for renewable projects to 0%-1.6% significantly lowers cost of capital for renewable projects (and energy costs/prices), and motivates the investment all of these economies/people/humanity needs.  Even if user/developer is using their own money for a revenue generating project, a cost of capital equal to the government bond rate is appropriate because renewable energy (with hydrogen production outlet to monetize surpluses) is risk free.

This policy is extremely valuable for assets with 50 year lifes such as solar and hydrogen equipment. It is especially valuable for hydrogen electrolysis from renewable energy because these are only used when sunny/windy, and so produce 4-5x less hydrogen than if they were used 24/7. High costs of capital tend to make any benefit from an asset that is more than 20 years away contribute close to 0 in present (investment time) value.

For home owners/builders, renewable energy generation/storage/hydrogen should be included, but so should non fossil fuel using heating/cooling/air-exchange/long-life-roofing-premiums and vehicle charging systems.  It is debatable whether other efficiency measures (extreme insulation/lighting) should be included in program as more solar generation/storage can be a more cost effective solution to eliminating carbon emissions.  Though dumb lighting/insulation decisions should be penalized.

For general economy purposes, government cost of capital loans should be given in the sectors of hydrogen manufacturing and renewable electrical transmission, electrical vehicle manufacturing (aside: EVs should not receive consumer subsidies. Emission reductions are based on miles driven not car ownership. Carbon taxes increasing the cost of alternatives provides the right consumer framework for choosing an EV), electrical and hydrogen powered machinery purchases and manufacturing that is normally powered by fossil fuels (tractors, mining equipment, blast furnaces). Loans in these sectors would be based on production delivered. Companies receiving these loans would effectively be receiving cash for projects they have already paid for. Repaying investors is a permitted use of the cash, but reasonable socially desired alternatives of expanding their successful production (which would result in another cash infusion loan after it is complete) is a recipe for rapid growth and international domination.

The cost of solar independent of electric grid
The cost of hooking up a 10kw solar system to the grid includes $3000-$5000 (brand name) for a grid tie inverter (theses have short warranties that could result in 4 replacements over 50 years).  The monthly fixed customer charge ($40 in Ontario and increasing) is $9600 over 20 years.  Connection fees to pole and meter $750.  There are additional cost/complications for using a battery system as backup to electric service (typically adds $2000 ($5k total) to grid inverter system).  There is a utility acceptance/inspection fee (and potentially capricious rules designed to increase expense) for connecting solar system to grid.  Using smaller AC inverters, while having main electrical loads (HVAC/electrolysis) DC is cheaper than a grid system, and DC allows 10%-14% more energy efficiency from solar and conversion avoidance.

The distribution network for an electric utility is somewhat complicated.  Typical wiring allows for a fixed number of 100 amp services on a branch (typically about 1000 or 500 200 amp services).  The size of service has to be limited due to possibility that all homes could draw full power simultaneously.  Many local power outages are the result of utilities cheating on this, or of equipment deteriorating such that 80% average power draw trips up the branch.

Limiting individual homes to 10kw solar is how the utility solves its distribution problem.  But the same parameters that allow a home to be self sufficient, allow 1000 homes on a microgrid/branch to also be self sufficient.  10MW solar distributed in any amount among 1000 homes will work on the same distribution wiring.  The homes with 40kw and 400amp service will never import 400amps, but all homes will prefer consumption when rates are cheapest (when it is sunny).

Solar value is usually miscalculated/misreported.  Modern solar panels will produce 66.9% of their new power after 49 years (0.8% degradation/year used in this paper, but there are reports that in cold climates solar degrades at only 0.2%/pa).  82.7% average production over 50 years.  15.09kwh is generated per 1w of solar per hour of sun.  The quick calculation of value for a 4.34 sun hour system (Toronto, 60 degree roof facing south.  See output for your location) is every 65 cents in installation cost is a 1c/kwh electricity cost, assuming 2% cost of capital and inflation cancel each other out.  A lower sloped roof produces more, and improves this simple calculation, but the model is wrong because winter electricity production is 2x-5x more valuable than summer production in Ontario (hotter climates may have opposite high value season).

A simple way to correct for high and low value energy, in the proposed 10kw solar system, all energy is used in the 3 low solar production (winter) months, and in the remaining 9 months, an average of 2.8 hours (28kwh/.day) is surplus, with 20 days of intense AC use, and February heat mostly drawing from January reserves.  Subtracting these surplus production hours from the 4.34 hour/day results in a "core"/essential energy production of 1.54 hours/day.  Each 23.2c/installed watt results in 1c/kwh core energy costs. 6.367MWh of surplus per year is enough to generate 141.5 kg of hydrogen (excludes any hydrogen production when winter is mild).  With 6kw of electrolyzers costing between $3k - $6k, and a 2% cost of capital, the (interest only) cost of hydrogen production is $0.42-$0.84/kg (with 50 year amortization(paying back principal) $0.63- $1.11... so a price per kg that fully pays back any loan funding electrolyzer purchases. Or 3.6% ROI if self funded.) with this surplus electricity.  If instead, 3.34 surplus sun hours (extra 0.54 hours) (10kwh/day electric consumption in 9 surplus months), it results in 169kg H2 production, and 35c-71c/kg (with amortization: 46c-95c/kg) production cost. Note that with a 90c/kg hydrogen cost (including amortization), the cost of diverting electricity away from hydrogen production is 2c/kwh (45kwh per kg production). This makes your electrolysis stack last longer, and you are paying the 90c towards the loan/capital costs of the hydrogen equipment whether or not you use it, less the small advantage of longer lifetime for the equipment. A productive example is that EV charging while sunny would cost you 2c/kwh. A different but relevant measure of the cost of diverting away electrolysis is the “opportunity cost” of not being able to sell it at a much higher price than cost. If you could use the energy to produce hydrogen that you can sell for $3.60/kg, then cranking up the AC costs you 8c/kwh. Current California retail price is $14/kg.

All of these costs are lower than natural gas at $9/mmbtu delivered (typical low retail), and without needing a $20/month customer charge gas utility hookup.  1kg of hydrogen burned in air generates 111k btu.  Does not require a chimney for 10%-15% efficiency boost, when burned with oxygen can generate 40% more btu, and can burn under water guaranteeing 100% heat delivery (another 10%-15% not wasted) .  Used in a fuel cell, with electricity run through a 300% efficient heat pump, and waste heat captured, a full 230% of the "burn btu" heat can be generated (with optional other electric use).  Thus a $2.30/kg cost of hydrogen is comparable to $9/mmbtu natural gas, given the optionality of efficient electric conversion.

The cost per solar watt depends significantly on whether it's new construction/design, DIY, or a turnkey system sold with warranties.  Savings from grid independence are significant.  Solar panels, wiring, and charge controllers should be $1/watt or less under new construction done at volume.  10kwh in Batteries can range from $2500- $5000, and a DIY used EV battery is at the low end of this range.  At the high end, as a professionally installed retrofit, $25k. (some of the costs eliminated from traditionally installed solar is large AC inverters, and batteries designed to tie in to large AC grid ties, though for retrofit to existing AC wired house, a tesla powerwall might add $5k to cost)  The cost of a 50 year roof isn't included but since it is less than double the cost of a 25 year roof, it pays for itself, especially with 2% cost of capital.

It is reasonable to take the $480/year hydro fee, and $240/year gas fee over the full 50 year life of the system.  Just a couple of years ago, the hydro fee was $360/year in Ontario.  Both are likely to keep increasing, but especially the hydro fee, likely faster than inflation.  Ontario has committed to wasting $24B on refurbishing its nuclear reactors, requiring a future 8c/kwh price/revenue increase.  Including the full 50 year costs though means that all low and high cost estimates are negative.  $24k in hydro fees saved, $12k in gas fees.  Even with Ontario tradition of having taxpayers (instead of rate payers) pay for its energy boondoggles, it's not reasonable to  deduct anything due to this as the tax payer cost "option" exists with a grid connection as well.  An additional savings related to connection fees and no boiler (better heat quality than furnace and close to solar design) are about $3k.   With 2 replacement boilers over 50 year period, it should be equivalent to water or hydrogen heating/tank replacement costs.  Batteries could need 4 replacements over 50 years, but costs should fall.  $2000 in 10 years, $1000 thereafter.

So the cost of the off grid solar system in Ontario is negative.  Save $4k-$19k with 0 energy costs, and revenue potential from hydrogen.  This excludes the savings from not paying for energy. 15kwh/day (same as this system) at 15c/kwh (will go up) for 50 years is saving $41k.

The value of hydrogen
1kg of hydrogen has the equivalent heat energy as 1 gallon of gasoline.  But when used in a fuel cell vehicle is 2x more efficient (more range) than a gasoline engine (slightly more with regenerative breaking), and so the real energy equivalent is 1kg H2 = 2 gallons of gasoline.

Hydrogen can be mixed into the gas pipeline network at a ratio up to 25% without damage to the gas network or any modifications to appliances that burn natural gas.  This is often recommended as a decarbonization process.  It would require immediate mandate that new appliances be dual-fuel, such that 12 years after the mandate is given, natural gas can be exterminated from use.  It is a relatively slow extermination process that also requires a massive 25% to 100% jump in hydrogen use when a switch is turned on in the future.

Hydrogen mixed with natural gas loses the special properties of hydrogen.  1kg H2 = 1/9th mmbtu natural gas, when mixed, but when not mixed, has at least 230% of the value because it can be converted to electricity much more efficiently than natural gas, then generate that much more heat, or even more useful electric applications.

Hydrogen competes with gasoline much better than it does with natural gas.  A carbon tax that brings gasoline to $6/gallon (just $2/gallon carbon tax in many jurisdictions) would permit hydrogen at $12/kg to compete.  A higher carbon tax applied to natural gas that would add $18/mmbtu ($3/gallon gasoline co2 equivalent), and make a pure hydrogen equivalent value of $6.90/kg. Assuming the gas utility/pipeline margin of $6/mmbtu would remain adequate, a 66c/kg distribution margin to utility/network would be charged.  This margin can be lower if hydrogen is used more because it has more value and more use than natural gas.  If the gas utility is delivering vehicle fuel (H2) to refilling stations replacing both oil and natural gas, a 40c-50c margin per kg would be higher than the differential between wholesale and retail gasoline ex taxes, and more fully use the gas network.

 Carbon taxes are essential for renewable energy adoption, because renewable energy adoption drives the price of fossil fuels towards 0, and that downward price move makes demand for FFs sticky.

An alternative to hydrogen production is trying to resell electricity to the utility.  The capital costs for 6kw of electrolyzers and a 10kw grid tie inverter are similar.  The 45kwh put into 1kg of hydrogen, after AC conversion losses are 40kwh that can be put into grid.  6MWh of annual surplus in Ontario would require an 8c/kwh credit to equal the $480 customer charge, and then a cash payment from the utility (not policy, and no enthusiasm from utility, as it would bankrupt nuclear dependence) of paying 12c/kwh to prosumers, Then $240 in annual revenue could be generated.  A $3/kg hydrogen profit would be $485 in annual revenue.  Requiring over 16c/kwh cash settled payment from utility to match.  Quebec, whose energy sector has not corruptly ruined the province with nuclear waste, has only a $12 monthly fee.  Just over 2c/kwh credit, would pay for the customer fee.  10c/kwh would match the hydrogen revenue.  Quebec electricity rates are 4c/kwh or less.  Generally, as more solar gets added on the continent, daytime electricity rates will come down close to 0 for summertime in Canada.

Hydrogen prices will come down with supply/demand balance in face of carbon tax.  Being paid $3-$6/kg initially will attract a lot of producers.  $3/kg will attract significant demand.  It can already be profitable below this price, and supply and demand adoption will drive down costs of both, and drive down the cost of hydrogen further.  If hydrogen is profitable at 10kw-50kw, its profitable at 50TW.


Upsizing to a 50kw solar home/farm
This section is about finding the right balance of batteries and electrolyzers, and the value of smoothing out solar production (either through tracking, or what is available to buildings, putting solar on 3 sides of building) in a direct to hydrogen production system.

In a simplifying process, I will ignore the slight annual production degradation from solar panels, offset by the fact that with 50kw no other heating system than electrolysis is needed, and reflecting the fact that first year(s) production will sell at a higher price than late years.

Adding an additional 40kw of solar panels to our system allows generating (over) 1kg of hydrogen per sun hour with “left over” 21.5kwh average daily household electric/cooling consumption (an increase over previous model in that it could be a farm, bigger house, or business).

A side issue that needs explaining is the effect of temperature on solar production.  Heat decreases production.  Cold increases it.  For Toronto, at 44 degrees south facing panels, the average net heat related losses are 8.4% (not reported so far in analysis) with gains in winter as much as 10%, and losses during peak sun and heat hours in summer of 20%.  This is an advantage for solar/hydrogen production in Canada, as more valuable winter energy is generated than "advertised", and generation is smoother during the long summer days.  Peak instantaneous generation will occur on a cold sunny day close to March equinox.  May exceed 50kw.  Designs that do not have enough electrolysis/battery/heat load to miss a few percent of peak power on a few hours of the year are acceptable, especially if other loads can be used at that time.

In determining the amount of electrolyzers with our 50kw system, 35kw of electrolyzers is a good start because a south facing array will produce over that amount for about 4 hours per day on average.  Halving the (fixed capital) cost of hydrogen production can be done by either halving the price, halving the cost of capital, or doubling the daily production hours (with half the capacity).  Its fairly rare for peak output to exceed 45kw, and rare still to do so for 4 consecutive hours, and so a starting battery capacity of 40kwh is appropriate. But reducing electrolyzers and batteries (done in following paragraphs) can lower costs of system.

The life of battery systems improves more than double when you double their capacity, because the system has slower charge/discharge rate, and all/most lithium-based battery technology has very little degradation when charged slowly (or not at all) above 75% state of charge, and discharged slowly below 25% state of charge.  It varies significantly by vendor, chemistry and technology advancements, but as a rule of thumb, an NMC-lithium battery rated for 3650 full cycles (to 20% deterioration), would do about 4x the half cycles, for 100% more energy life, and then double the energy life by using slower charge/discharge rates around the extremes.


NMC cycle life as affected by depth of discharge  
If batteries cost 1/4 per kwh ($250) that of electrolyzers per kw ($1000), then replacing 10kw of the electrolysis (down to 25kw total) with another 40kwh of batteries (now 80kwh) has the same overall capital costs.  Yet gains are achieved by having a greater portion of the day be over 25kw (50%) solar production (using full electrolyzer capacity), and the portion of solar day over 25kw production goes up to 6 hours on a decent day.  Gains are also achieved by a maximum battery charge rate of 25/80, and that maximum occurring at solar noon when batteries would be at most at 50% charge, and the maximum charge rate is actually very rare cold spring day.  As long as there are very few days where the 6 hour peak doesn't average over 38.6kw, then 80kwh of batteries are sufficient to capture all energy.  Cloudy days and winter days that produce under 3.1 full sun hours will under-utilize the battery.  Because batteries don't last as long as electrolyzers, and there is some inefficiency charging/discharging them, the cost differential has to in fact be more than 4x in order to replace 1kw electrolysis with 4kwh battery, even when it is both sufficient to capture peak energy and charges 100%+ of its capacity each day.

Without electrolysis (or other useful/monetizable dump), large battery systems are wasted because there is no dump available to use their energy storage.  It is energy discharge that determines battery value.

The 2nd improvement we can make is to shift 20kw of solar panels to east and west sides (10kw each).  This creates 2 daily smaller power peaks (about 35kw) instead of a single one (50kw).  This reduces total annual power output, but due to smoother daily production allows us to utilize electrolyzers more.

Since we no longer care about winter production maximization, but instead want annual maximization, and we can use more roof space, A 44 degree south face is optimal and provides 4.61 sun hours/day in Toronto.  70 degree east/west angles can be placed along a wall, and capture 3.05 sun hours/day, but with good performance at sunrise/sunset.  This reduces average daily production (compared to 50kw south facing) from 235kwh to 199.3kwh (not including the 8% temperature related loss). This 15% power drop can be profitable if it enables more than 15% reduction in electrolyzer and battery costs.

The 3 sided solar arrangements produce 2 peaks at 9am and 3pm.  The 50% electrolyzer strategy (to capture good 6 hour days) now requires 20kw electrolysis.  The battery strategy that captures all energy if production averages still 77% of south capacity (+10% of east OR west capacity) over those 6 hours (23.1+1 kw) means just 24kwh in batteries.  This is a 20% reduction in electrolyzers and 70% reduction in batteries.  Production is smooth enough to achieve the full 10 hour average electrolyzer utilization.  This is 14kwh extra battery capacity over the 10kwh needed for general home/property non-daylight use.  This arrangement is not obviously better than the equivalent capital allocated (using 4:1 ratio) of an extra 3.5kw electrolysis (instead of 14kwh extra battery). Though production will frequently exceed 23.5kw, the (smaller)10kwh battery allows a higher peak 6 hour average of 25.1kw (1kw higher).  On the other hand, when power is between 20kw and 24.1kw, the larger battery allows for banking the power to extend electrolyzer use later.  If the 2 options are in fact equivalent, choosing between adding either more electrolyzer capacity or batteries, its likely that batteries are a better choice:  Battery deterioration will mean more headroom for home/property non-daylight energy use, and/or a longer time for battery replacement.  Days where a battery does not fully charge means a day longer of life for the battery.

So our final system adds 40kw solar, 14kw electrolysis and 14kwh of batteries.at a cost of $57.5k.  Our property uses 15kwh of non-heat electricity per day.  Heat is a free byproduct of electrolysis (about 50kwh produced on average 3 winter sun hours) , and does not require heat pumps saving $5k or so.  Final bill $52.5k (negative cost of initial system is irrelevant to decision for this higher production alternative).  The system will produce 46.26MWh in first year from south facing panels including thermal losses on solar cells from hotter days. 20.824MWh for first year sides power. Less 5.475MWh annual household consumption and the 82.7% production factor over 50 years gives an average annual surplus of  50.95MWh.  Enough to make 1132kg of hydrogen per year.  At a 2% cost of capital ($1050/yr), a cost of 92c/kg of hydrogen.  Backing out the $17.5k in extra electrolysis equipment, though adding back the $5k in heating equipment savings, $40k at 2% capital cost is $800/year.  If 50.95MWh surplus can be monetized at 1.57c./kwh there is break even.

This section used costs already achievable/beaten by utility/large scale projects, though without the benefits of solar tracking hardware which is worth it because it provides all day power smoothing with maximum utilization of panels, and so less batteries and design.  The key takeaway should be that cost of capital can determine/provide ridiculously low costs for renewable power and hydrogen production.

Its worth noting that the larger system results in slightly higher cost of hydrogen production than the smaller one because there is more surplus (to hydrogen) energy on the larger system and so more electrolyzers and batteries needed to convert it all. If the costs of small systems can be equal-scaled to that of larger systems, then distributed solar/hydrogen-production is that much more attractive.

Modelling the cost of batteries and electrolysis
A 10000 cycle life battery adds a 1c/kwh discharge cost per $100/kwh of battery cost to the charging cost of the battery. Ignoring capital costs can be justified if it is low through supported policy, and balancing it with residual value (still has some charging capacity) of the battery at the end. A $300/kwh battery means a 3c/kwh battery cost. The importance of sizing the battery such that 40%-60% of its capacity reflects typical non daytime needs (ie. 10 kwh battery for 4-6kwh night electrical use) is to maximize its planned cycle life by targeting usual discharge rates. The symbiosis with hydrogen electrolysis is ensuring multiple partial cycles per day around discharging ahead of predictable solar charging peaks.

The planned battery costs are actually higher than real costs because the replacement of end of life batteries will be much cheaper than at the current ones' time of purchase. So real costs are based on the expected cost/kwh of the replacement battery, and using the battery cycles for electrolysis may bring life from 30 years to 10 years, but batteries have time-based degradation and many cycles/day over 10 years has much better value than 30 year hoped life.

So, a 10kw solar system costing $13k (2c/kwh) with $3000 for 10kwh batteries, if all electrical output is used, and most of it used during daytime costs an average under 3.5c/kwh. As noted in above sections, this is not the right model. Instead, the above system provides 5621kwh year of “essential” resilient energy (15.4kwh/day average) costing 6.9c/kwh, with the surplus of 6467kwh per year costing 0, though any monetization/use of the surplus reduces the “real” energy costs. These figures include “free” heating needs in Canada.

For home/business production, the returns from hydrogen do not need to match the net/perfect 3.5c/kwh from full generation utilization. If the government will “reimburse” you for the $3000-$6000 in electrolyzer purchases with a 2% or less loan once you demonstrate that they are part of a sensible/functional system, what is enough to make over this amount? I'd suggest that a 50% target over this cost is worthwhile, and reimburses you for the risk of long term hydrogen pricing. At $1/watt electrolyzer costs this would make a sale price of $1.66/kg reach this profit target.

Gas utlity competition to electric utilities
With a moderate $18/mmbtu carbon tax, gas utlitilies would be incentivized to switch to hydrogen.as their fuel of choice.  It has already been discussed that a pure hydrogen network has far more value than hydrogen mixed in with natural gas.

Gas utilities delivering  hydrogen at $4/kg would let gas customers get their electricity from there through a fuel cell.  Each kg provides 18kwh of electricity and 15kwh of heat.  The 40% heat corresponds to the usual fraction devoted to water heating, and so with "free"/included water heating, $4/kg corresponds to a 12c/kwh electricity price (excluding fuel cell purchase cost).  When hydrogen is produced under $1/kg for later self consumption, if the heat is useful it costs under 3c/kwh (cheaper than batteries most of the time), and if the heat is not useful under 5c/kwh. A gas utility taking 3c/cubic meter profit on delivery of hydrogen purchased at the $1.66/kg above cost, means delivering to customers $2/kg hydrogen, and 6c/kwh domestic energy value.

Gas utilities do not enjoy the same monopoly abuse as electric utilities.  Customers of gas companies have the alternative of having propane or heating oil delivered, and such delivery businesses can start up quickly with modest capital.  While the fossil fuel industry faces an existential threat as a result of their existential threat to humanity, the gas distribution business does not depend on carbon emitting gases.  Electric utilities is their current biggest customer, and when they lose natural gas electric generators as customers, the network may be underutilized.

What makes hydrogen a valuable energy carrier is that it is permissionless.  Electric utilities are the only buyers of electricity generators, and can slow walk the transition to renewables.  Electric transmission has expensive limited capacity with losses, expensive conversion between current types and voltages, and balancing demand with its other/legacy supply interests.

Gas transmission doubles as storage (Germany has 3 months (ie winter) of energy storage in its pipelines), and has simple interconnections without conversion losses or power loss over distance.  Hydrogen though less power dense than natural gas has the advantage of being able to flow faster through even small pipes, and doesn't raise environmental concerns.  Furthermore, Quadrupling pipe capacity (doubling diameter) increases cost by 40%, while quadrupling electric transmission nearly quadruples cost.

2004 Canadian estimate of gas transmission and delivery cost put a cost per cubic meter-km at $0.00003.  This is 15-25 times less (depending on whether value of waste heat is included) than my estimate of a transcanada electric trade route would cost (though with 10% cost of capital, and unclear how the linked estimate considered cost of capital), excluding electric transmission losses. Fiber Reinforced Plastic pipe (10cm diameter) suitable to hydrogen transport at 240bar costs $50/meter to install

The storage function of the gas network has very high value.  If a 300 atm tank costs $1/liter and lasts 120 months, then its storage cost per cubic meter-month is close to 3c. (A run of 10cm FRP buried in back yard costs a bit less)  The same cost as transporting it 1000km to a customer.  If a small hydrogen producer (that does not consume) is charged $1000 to connect to the gas grid without a monthly customer charge, then the arrangement is far more cost effective than having tanks on site, and paying for pickup services.  The gas network furthermore requires less compression level (up to 240 bar, but based on how full network actually is)

Electric vehicles can be powered by battery or hydrogen.  For personal vehicles that can be charged slowly at home, and driven up to double the national average range, batteries offer a more efficient energy cost over hydrogen.  But for heavy vehicles and airplanes, hydrogen offers more range, and in the case of any commercial vehicle where refueling time is extra operator cost and lost revenue, pipeline supplied hydrogen refueling stations can supply energy over 2x cheaper than fast charge electric stations because they can refuel more vehicles with more energy with cheaper infrastructure, and they are not dependent of time of day electricity prices.

Industrial heat is the largest consumer of energy/carbon (surpassing electricity generation and transportation) in the US, and generally achieved through combustion rather than electric heat.  Industrial sites are usually high electric consumers too, and can get huge benefits from solar self generation with surpluses going to hydrogen production.  The value of self generated hydrogen for combustion is comparable to natural gas without large volume discount or discounted natural gas with a small ($6/mmbtu) carbon tax.  Hydrogen for electric generation is 130% more valuable.  Pipeline access to hydrogen and conversion of some heat to electric would reduce daytime energy costs, and provide power flexibility/options for additonal shifts, power requirements higher than solar generation, fuel for shipping vehicles.  So power and heat needs can be better provided to industry by the combination of self generation and hydrogen export/import from sites.

Electric utility challenges to overcome
The existance/threat of a hydrogen delivery network is a significant threat to electric utility monopoly power, and potentially its existence if it does not adopt competitive consumer friendly pricing and other policies.

Consumer generation of electricity eliminates the cost of long distance transmission, and through time shifting, reduces the need for baseload generator services.  Not building more transmission is a saving that can lower prices.  Paying cash for consumer generation can encourage it and use its existing infrastructure to profit with no investment on its part.  Disinvesting from legacy power agreements the utility does not own, can reduce use of high priced legacy energy.

Renewable projects that electric utility supports through tranmission lines should have 4 to 12 hours of storage to fully utilize the transmission lines.  This is worse (more expensive) than the optimum 1 hour storage + hydrogen, but it is less expensive than legacy baseload, and reduces time shifting transition from consumers.  The consumer base that will stay with electric utilities will do so out of habit, and not pursuing savings investments of their own.

The electric utility can be the utility that funds/launches the hydrogen distribution network in order to preempt an independent or gas utility owned one.  This pairs well with providing fair/attractive prices to consumer generators, or letting large scale renewable plants consolidate electricity surpluses into hydrogen production facilities such that it can offer fixed power purchase agreements, or guarantee consumer generator minimum prices of say 3c/kwh.  3c/kwh price adds $1.35/kg to hydrogen prices, but it may be a policy that saves electric utility model, while still providing reasonable priced hydrogen.  Without hydrogen generation, there is no way to guarantee a minimum/fixed price to generators, unless supply is limited and energy decarbonization fails.  Neighbourhood level electrolysis is an option/necessity to deal with local surplus power.

The 2% or lower capital cost financing proposal is available to electric utilities as well.  Colocating energy assets (solar and storage) owned by utilities but rented/consumed from by consumers is a way to leverage the best efficiency that occurs from site produced/stored energy, leveraging salaried expert installers and large scale shipping volumes, and providing consumers with cheaper energy without investment, while not requiring land/support investments by utility. Consumers can be made happy just by giving them a small monthly bill reduction even if most of the profits are kept by the utility.

It's this last opportunity that provides the main opportunity for electric monopoly utilities to maintain control over consumers while profiting from the value they provide at no consumer risk.  For instance, a 10kw solar + storage system can be accompanied with an $X monthly discount equal to monthly charge + 15kwh/day consumption with the consumer benefit of uninterruptible power supply.  The same deal can be offered even if much more than 10kw solar is installed on site.

With a 100% renewable purpose, hydrogen (or some other industrial "energy dump") is necessary to add value/subsidize the needed surplus electric production seasonality.


Stranding assets as the cost of green energy transition
“Pro business” conservative/republican politicians reward the rich and powerful from status quo state by giving them more money and power. It is an anti-market, anti-disruption, anti-innovation stance that suppresses investment and opportunity.

Green New Deal proponents who propose government centered spending to spur energy transition, though, are wasting resources. The 2 “free” policies focused in this and previous papers – A carbon tax and dividend, and offering cheap cost of capital for completed green projects. An important 3rd “free” policy is to allow cancelling/renegotiating fossil fuel related purchase contracts. Legacy generators can, in the short term, be paid for standby capacity. Such cancellations would help utilities accelerate renewable investments and plant closures, in coordination with carbon taxes. (A great 4th free policy led by California is mandating solar production from new buildings)

The 20th century fossil fuel energy system was not built with government money. New energy does not need to be either, and politicians promising to control that money is not a recipe for providing value in the energy transition. Bernie's climate plan is best for shunning new nuclear energy, but Andrew Yang's high carbon taxes (helping fund freedom dividend) makes his climate plan the best even if he has poor expertise in the climate and energy area. Andrew Scheer (canada)'s pure evil climate plan is the worst possible central planning corruption: He will hand free money to polluters so that they can research, using the funds, how best to exterminate themselves. Just as the EPA has been transformed into a pro-polluter agency, any government managed discretionary program is likely to dangerously deviate from purpose/mission over time.

The entire cost of energy transition will be borne by the stranded assets that climate terrorists created and those evil and stupid enough to still own/fund. A common zombification is that “the successful deserve to have all of their assets protected from disruption”, because their initial success was due to some favourable balance of hard work, talent, luck and starting position.

Though the previous 2 sections dealt with a path to maintaining relevance for energy monopolies, there are paths to life without them.

Independent electric and hydrogen microgrids
The first problem with utilities is their $12-$40 fixed customer charges. They need a fixed charge because of billing, meter reading, chasing and dealing with non payments, and funding their bureaucracy. But high monthly charges are entirely the result of past corruption/failures and an effort to pretend/disguise that per kwh rates are reasonably unshocking to “boiled frog” captive customers.

Cryptocurrency/blockchain/smart contracts similar to the bitcoin lightning network can allow settling energy transfers seconds or minutes after they are exchanged. Easy to use computer interfaces can manage it all, and it becomes worth dealing with to avoid the monthly charges.

A peer to peer (neighbour to neighbour) electricity network can be set up in urban environments as follows:

  • Standardize on 120V DC current (or voltage of common AC current in your region)
  • #2 gauge wire(s) costs $15-$20/meter. Will carry 9.6kw 56 meters with 5% power loss. Doubling/halving distance or current will double/halve those losses.
  • Neighbour that would need to receive net power from a solar generating neighbour would offer to pay for connecting the 2 points/homes.
  • A meter would be placed at generator's home to tabulate net energy transferred. Connecting/paying neighbour may choose any wire size they wish but pays for energy before line losses.
  • Base pricing can be model based around seasonality, weather, time of day, where guaranteed production percentages are made available to importer, and guaranteed minimum purchases are promissed to generator. A mechanism to adjust access and minimums with long term notice. Outside of those parameters, market (bid-ask based) rates would apply. Computers on each party's behalf manages trade based on consumption patterns and each side's battery capacities/charge rates.
  • Solar generating buildings may receive connections in 4-8 directions to neighbours.
  • A net importing neighbour can receive a connection on the other side of his main connection. That new connection becomes the first importer's “customer”. First importer can set a fixed profit spread on 2nd importer's consumption, and to handle times of scarcity, the first importer can guarantee that at least 50% of energy transferred at 2nd importer's bid prices are passed through.
  • 3rd importers would connect to 2nd. As customer of that 2nd.
  • As strings of connections get longer, anyone part of that string will see the proposition of installing their own solar generation more attractive, as it supports lengthening the string length and profiting from sharing.
  • When “strings” emanating out of generators meet/become adjacent, it is to both of those neighbours advantages to connect to each other in order to diversify their supply routes, including the potential to profit from energy trade in the opposite direction of their main/initial supply source. Which side pays and which side controls the meter is a matter of negotiation, but the side with the “official” meter paying for it as a minimum offer is a good starting point. If both sides want the meter, then the negotiation becomes balancing paying for the wire and smart contract terms for trade/spreads between the two points. Adjacent neighbours may wish to help subsidize/assist connecting “strings”, or offer to jump over obstinate neighbour, as they also benefit from supply diversification.
  • An urban solar production facility, under this scheme, has the capacity to deliver 20kwh over 4 hours to points 112m from any of its edges at 5% loss. 224m at 10% loss. Much more power delivered to closer points (allowing them to share with further points). Any partial solar production or batteries along the route extends the delivery amount and time window and practical distance. Where an urban/suburban environment can be modelled as a chess board with 20m-side squares (400 sq.meter properties), 8 25 square chessboards can be placed around a large generator, each consuming an average 5kw. 1MW of solar distributed through the super chessboard. This power is almost sufficient for Canadian winters, but its massive in summer. More generation throughout the square is useful, as are fuel cells to meet winter peaks/resiliency at high rates for that power. The exact same dynamics that have been described for single home resiliency apply to “chessboard peer to peer” microgrid. Surpluses need to be diverted to electrolysis.
  • Incumbent utility can participate in microgrid, offering no-monthly fee crypto settled connections (that are likely already in place) to any single member of the chessboard, and doing bidirectional trade with that/those member(s) and serving far away generators and customers to help smooth out any surpluses/deficits in a wider region.

The 5% edge of chessboard losses are less than utiltiy provided AC distribution which is 6%, and converting to AC at the last possible consumption point allows 14% more solar or battery/fuel cell power to not get wasted. These wastes are especially apparent with distributed “virtual” battery systems where AC conversion back and forth is a 10%-14% hit each time.

The microgrid may have difficulty supporting large apartment buildings (can also use hydrogen pipes within building for heat/unit level fuel cells) and industrial/commercial users users. Any resilient energy system has to include surplus moments and so access by the electric microgrid to a hydrogen network is needed to dump surplus energy. If automakers' claims of 36kw fuel cell currently costing $2000-$3000 are true, then there is a substantial profit opportunity from even small homes (but obviously from industry/apartment). The top 200 hourly market prices each year (cold december/january nights) could average 30c/kwh. A 36kw fuel cell finding 20c/kwh profit over just 100 hours each year would make $720/yr. More than enough to pay for the fuel cell. There may also be many situations where heat is needed and electricity not by the fuel cell owner, and so selling that electricity from a fuel cell would have competitive advantage over battery discharged energy. The fuel cell provided electricity may also come from a FCEV (car) with vehicle to grid connections, though a car cannot burst as much energy into the grid as quickly.

For sparse rural properties, not choosing a self sufficient solar production is retarded. Not having space/solar access is the only excuse not to go solar. They need access to a hydrogen network mostly in order to dump all of the energy they are able to generate. Not farming solar energy on a large property is also retarded. Even if they do not need to consume hydrogen, a pipeline somewhat near them likely connects urban, industrial areas or distant wind farms. Access for supply reasons though opens up demand opportunities.

10cm 240atm FRP pipeline is extremely cheap to install ($50/m) as it comes in mile long spools and do not need welding for longer runs. A small crew can install miles per day. It delivers up to 33MW of energy sustainably all along the pipeline, 18MW of which can be converted to electricity. It is bendable in a relatively tight circle (2-3m radius), and suitable for buried compressed storage, or coiled at bottom and sides of a circular pool. Technically, it might be feasible to have the same peer to peer hydrogen network as the electrical system, but a neighbour is not particularly likely to need hydrogen if you can provide them with electricity, and the cheap installation costs of spooled FRP come from deploying long lengths.

That one entity would own a long segment of pipe does not mean that crypto transactions settled near immediately aren't the right way to deal with hydrogen flow, and does not preclude different ownership of different pipe segments. There are 3 special issues with a decentralized hydrogen network transactions: 1. Depositing into the network does not imply a sale has been made. 2. The losses in the network while lower than electric wires or batteries, exist, and if there is no broker guaranteeing a price, need to be attributed to the seller/depositor. 3. Segment owner refuses to quickly repair a leak in their segment, compromizing an entire line value and potentially the contents therein. Addressing the 2nd issue first...

Losses in a hydrogen pipeline (of constant size) are a linear function of pressure. At 103bar pressure, 10.1cm diameter pipe from fiberspar will hold 8252 compressed liters per linear meter. 220bar capacity is 1.45kg/m. Losses would be 0.005 liters/day. 6/10000th %/day in losses. The unit losses double at double the pressure, and halve at half the pressure, but the percentage is constant. These losses are very low but need accounting.

To the 3rd issue, breakage and repair, chain of custody through pipes must be tracked, a bond or insurance by segment holder is required to cover potential loss of gas, and a service contract responsibility must exist for the whole line. The only practical solution is that the entire line of pipe is an organization whose shareholders are the segment holders. Something like 1 share per meter. The organization collects revenue from network withdrawers, and funds a budget for inspections, maintenance and repairs. Blockchain helps shareholder accounting too.

As to the billing issue, a single entity for dispensing or deposit transactions is a major simplifier. There are several mechanisms to address the deposit/withdrawal lag. 3 day rolling futures contracts where depositor and withdrawer have 3 days to complete their commitments or buy back missing portion of contract at a penalty. Another method would allow the pipeline to be used as storage, letting a depositor place a reserve (sale) price, and to discourage setting too high of a reserve price the depositor pays a storage fee for the time until their price (which can be modified) is met. The storage fee would rise as pressure in (how full) the pipeline rises. Previously deposited hydrogen with a reserve price can be switched to meet “futures” delivery commitments, and market purchased deposited hydrogen can offset “futures” purchase commitments. Derivatives that allows speculators to provide producers and consumers with fixed seasonal prices should also be part of trading platform.

Network compensation is a fee per cubic meter delivered as well as the above storage fee. Alberta has one of the lowest gas delivery fees at 3c/cubic meter. That seems very high long term for hydrogen. For one, hydrogen has the potential to replace not only traditional natural gas uses, but also fuel the entire transportation sector. It is less energy dense than natural gas, and so more is required for same purpose. A 1.5c/cubic meter delivery fee would be equivalent to 1c/kwh from electrical fuel cell output. This still seems high as daily turnover of the pipeline would return 27c/day/meter giving a payback on (20-50 year life) pipeline cost of under 200 days. Though there is an operational budget to fund. The storage fee would be paid in hydrogen, transferring a portion of reserve-priced deposits to the network organization each 1 to 4 hour period, and offered for immediate sale at a matched rate to lowest offer price.. These fees would more than cover the diffusivity losses in the network. 10x the loss rate for up to 100 bar pressure, 500x at 200 bar, and 2000x over 230 bar. These are percentage rates of 0.006%/day, 0.3%/day, 1.2%/day. The main reason for the storage fee is that prices are likely to be highest in winter (though southern US has solar surplus potential in winter and north/south pipelines are allowed), and spike on a specifically bad weather winter day, and depositing in late spring hoping to get winter emergency prices would mean 8 months of storage. As the network gets full, incentives to clear it out need be applied. Payment in hydrogen does lessen the cost as the percentages get applied to a declining balance. The storage fee allows significant reductions in delivery fee, but also encourages alternative (self) storage arrangements as pipeline fills up.

Interfaces to network clients (both suppliers and consumers) are paid for by clients. Standards for meters, gas purity of supply systems/equipment, consistent dispensing interfaces, and roadside point of sale systems, and perhaps valve checkpoints on each large interface, all need to be enforced by the hydrogen network, but ideally, through a single global standard. Supply interfaces could be responsible for adding pressure into the pipeline, and demand interfaces that need very high flow (or pressure) would be responsible for sucking pressure out. The normal pipeline structure to maximize flow rates involves repeating compressors along the length. If we assume many suppliers though, there will be enough pressurization, vehicles can store up to 700 (900 in future) bar even if they don't draw much from the pipeline, so sucking may be part of required refueling dispensers anyway. As the hydrogen economy successfully overtakes much of the regional energy mix, then it will make financial sense for pipelines to invest in improving flow rates (and invest in larger pipes).


Direct to hydrogen “utility scale” renewable projects
Avoiding the cost of high voltage AC conversion and transmission lines is the next step in drastic cuts to renewable energy project costs, though cost of capital policies reducing it from 8% to 2% is a 50% cost reduction, and 50 year PPAs another 10%-20%). A record low US Power Purchase Agreement of 2.175c/kwh was set in the fairly north state of Idaho. The justification for what is likely 1c/kwh lower than a site at that lattitude would price a contract is that it is taking over the electric substations of a decommissioned coal plant. By all means replace all legacy generating assets already on the grid with large solar+battery systems that reuse those expensive substations, and the limited transmission lines that feed cities.

The reason we have Time-Of-Use pricing on our grids, despite legacy generation's baseload/constant power output, is to shift/equalize use of our transmission lines. Daytime rates are higher than night. Reversing TOU rates such that they are low when it is sunny not only requires high solar penetration (40%+), it requires high distributed/urban solar production that does not use limited transmission capacity. As shown earlier, a hydrogen production outlet for renewable energy guarantees a price floor for the electricity production. A sale price of $1.80/kg H2 is equivalent to 4c/kwh electricity sale. If the amortized cost of electrolysis is only 45c/kwh, then even if hydrogen prices drop to 45c/kg, a guaranteed price floor of 1c/kwh still pays for the electrolyzers, and you can refuel your quiet hydrogen powered speed boat for 22c/gallon-gasoline-equivalent energy instead of selling it.

The power of having a price floor for energy projects, and a capital cost policy that provides near upfront cash repayment of most of the project cost, is that a PPA agreement is no longer necessary for financing renewable projects. The critical benefit is that renewable project plans no longer need permission from climate terrorist supporting utilities with business model ties reliant on climate terrorists, and regulated by climate terrorist supporting politicians in order to install all the projects they want. Not being limited by transmission availability further expands project locations.

Going direct to hydrogen is a different proposition for solar vs wind. For solar, tracking is essential to have a smooth power curve during a long day and fully utilize electrolyzers. A small battery may still be useful for modulating through cloud cover, but the simplicity of direct connections to solar may reduce electronics and intermediate losses. $1/watt solar and range of $500-$1000/kw electrolysis-45kwh/kg. 8 sun hours (southern US) tracked/day 82.7% of production average over 50 years = 120.7kwh/kw - 0.83c/kwh electricity cost with capital cost = inflation. Produces 53.66kg/year per kw solar+electrolysis. At 3.2% amortized (2% cost of capital + 1st year principal over 50 years) repayment = 29.8-59.6c/kg of hydrogen for total (+ 0.83c/kwh * 45kwh/kg) of 67c-96.8c/kg (ignoring heat degradation). The same setup in Toronto results in 1.18c/kwh, 37.56kg/year, 42.6c-85.1c/kg electrolysis amortization, and total hydrogen cost of 95.8c- $1.38/kg. Residential/business systems do better than these prices because they use the electricity/heat (and land), and the number of electrolyzers is allocated to handle just electric surpluses.

Wind power is highly variable. A turbine rated at 8MW for 10m/s wind will produce 1MW at 5m/s. The electric generators coupled to turbines almost all cut off at 12m/s or 10m/s due to challenges of having to transmit higher power. The same blade design that produces 8MW at 10m/s, would produce 64MW at 20m/s.

A 8MW turbine operating at 40% capacity produces on average 3.2MW. If that 40% were produced as 9.6 hours of 10m/s wind and the rest of the day, no wind, then 3.2MW of electrolysis would need to be paired with 46.1mwh of batteries, which happen to luckily take exactly the remaining non-producing part of day to discharge into electrolysis. 100% electrolyzer use. Constant 3.2MW production also has 100% electrolyzer use with 0 batteries. 12 hours of 6.4MW average with 12 hours of 0 needs “only” 38.4mwh of batteries. I'll use this latter size still hoping for 100% utilization at 3.2MW. 30 year lifetime including battery as it has slow charge/discharge rates. $1.3/w for turbine, $250/kwh for batteries, $500-$1000/kw for electrolyzers: $10.4M turbine, $9.6M batteries, $1.6M-$3.2M electrolysis = $21.6M to $23.2M total, amortized payment (4.5%/year) of $1M/year (near higher end of $1/w electrolysis cost). 28.03gwh/year production = 3.56c/kwh 24/7 baseload power (if no electrolysis equipment, 3.1c/kwh baseload). 622.9 tons of hydrogen produced. Cost = $1.605/kg.

The 4:1 cost ratio of electrolyzer kw to battery kwh cost mentioned at beginning of paper is important, and in the above scenario, there is way too much battery. 6MW of electrolyzers with a cheaper turbine whose generator tops out at 6MW, and costs $1/watt brings cost down to $11M-$14M. If capacity factor goes down to 30% or 2.4MW average, amortized costs are $495k-$630k/year, hydrogen production 467.2 tons. $1.06-$1.35/kg

Offshore wind has higher capacity factors due to high sustained winds, and massive efficient turbines that may be shipped/erected there. If removing all electrical connections to shore brings costs down to $2-$3/watt and allows floating hydrogen production anywhere on the ocean, then either trans-ocean pipelines filled by strings of wind turbines are possible, or ship refuelling at sea, or tanker pickup of hydrogen in hops. With 70% capacity factor, and full 8MW of electrolyzers, for turbine costing $16M-$24M, and electrolysis $4M-$8M for total of $20M-$32M. 1090 tons hydrogen/year produced at cost of $900k-$1.44M. $0.825-$1.30/kg cost.

While large scale renewable hydrogen projects described here do not beat co-purposed small scale locations (described earlier), the small scale costs are aspirational based on these large scale costs. The same case for locating renewables that reuse AC transmission infrastruture left behind by exterminated legacy power generation, direct-to-hydrogen projects located near exterminated gas supply pipelines to get free

The other case for hydrogen in mass scale renewable energy projects is as a hybrid with AC power transmission and batteries. Very similar to the small battery home scenario, the battery serves as an arbitrage device to sell power at night, and hydrogen production serves as a price floor for surplus generated electricity. For traditional renewable only, and quite common recently, renewable + battery, this reduces costs of transmission and battery size, and reduces electrolyzer sizes from this section, gaining the economic benefits of producing hydrogen with 0 cost surplus energy from earlier sections, and supporting 100% renewable energy that is 100% competition risk free.

Ramping up to green hydrogen economy
There is a cost decrease, supply, demand, infrastructure feedback cycle for hydrogen, the most important of which is cost relative to climate destroying alternatives. Batteries are not really a competitor to hydrogen because hydrogen has unlimited storage potential at close to $1/watt-hour, more efficient and cheaper than electric transmission at a distance, and sufficient energy density to power any transportation method including rockets. Coal is the easiest fossil fuel to exterminate, and is happening at a reasonable though still slower than ideal pace. Gasoline is 2nd, natural gas last.

A carbon tax in the US is the most important policy for fighting climate change, and accelerating energy transition, because it is the biggest international market, and border carbon adjustments (similar to tarrifs) would compel countries already more publicly supportive of fighting global warming to adopt (with business support) carbon accounting and retaliatory carbon pricing for their international trade. It would be contagious for all of those other countries' trading partners. Political acceptance of a carbon tax is easy if the revenue is rebated as a cash dividend to tax payers, and even easier if it partially funds an even larger freedom dividend/UBI.

A carbon policy that creates a path to $6/gallon gasoline is the best policy to spur battery and hydrogen electric vehicles. $12/kg hydrogen parity means $6-$8/kg is a significant incentive to switch. Commercial vehicles are the most important to oil extermination, because it is not owning an EV that reduces emissions, it is driving one long distances. Electric buses in China have displaced much more oil use than global electric cars have. As long as a carbon tax ramp up leads to a somewhat near term $6/gallon gasoline price, oil will get exterminated as long term ownership of, especially commercial vehicles, will be too costly.

The hydrogen economy is more suited to cold latitudes as it provides a heat bonus to electric generation, and vehicle range is not compromised by cold compared to batteries. Just as importantly, cold latitudes have massive solar energy surpluses in summer when meeting winter energy needs with renewables.

Because too many people are stupid and easily deceived, an even easier politically supportable policy than giving people more cash back than carbon taxes would collect is low interest loans for successfully deployed green economy projects. For renewable energy projects, only those that include a path to hydrogen production justify risk free interest loans, because it is only those projects that guarantee a price floor for energy, and guarantee use/marketability (distant exports an option) of all energy produced. These loans cost society/government nothing, generate jobs and tax revenue, and support national industrial policy for green energy dominance/competitiveness. Whenever no one left in the world can justify a larger speedboat, and earth consumption of hydrogen no longer justifies production increases, we can use hydrogen to exploit/colonize space.

2% or lower capital costs for hydrogen electrolysis and delivery systems makes electrolysis today cheaper than hydrogen production through fossil fuels (when natural gas costs $4/mmbtu or higher, which best wholesale requires only $1-$2/mmbtu carbon tax to exceed fossil costs everywhere), where the hydrogen production cost bar is set at $1.50/kg. Hydrogen is already part of the global economy. Over 82M tonnes in 2018. 60% of which is non-fossil-fuel-refining related. With 2 hours/day surplus solar, and 6 hour/day electrolyzer use, this production would support 1.68GW of electrolyzers (800 times 2018 sales), and 5.05GW (~5% of global production) of solar. Electrolysis has the advantage of producing hydrogen closer to where it is needed for chemicals, including smaller batches close to consumption centers of those chemicals. So in addition to saving costs on natural gas shipping, it saves on shipping cost/time to customers.

While an $18/mmbtu carbon tax on natural gas benefits all of society when the tax proceeds are redistributed as a cash dividend, there exists the regulatory alternative of forcing of natural gas networks to include up to 25% of volume as hydrogen in their pipelines. Such a ratio does not require any device using natural gas to be changed. An equivalent to $18/mmbtu carbon tax (where pipelines are encouraged to blend green hydrogen without being forced to) is a $2/kg premium over the natural gas energy equivalent of hydrogen of $1/kg per $9/mmbtu-nat-gas. Forcing/enticing natural gas pipelines/distributions to pay $3/kg for any green hydrogen brought close to their network would bring a gold rush in electrolysis. Over 100M tonnes of green hydrogen production/year would be supported by this policy. With previous parameters, another 2.05GW of electrolysis equipment, and 6.15GW solar is supported by these policies.

The above policies are sufficient to rapidly increase production/demand for electrolysis, and substantially decrease costs as a result of their learning and scale. This process can happen intensely one region at a time. The resulting cost and price reduction of hydrogen will drive demand, and the supply of and research into fuel cells and their applications. It would cut any previous hydrogen roadmap timeline in half.

There are 3 interesting fuel cell sizes, though I'm not sure the “packaging” matters that much. Bicycle (50w-200w), fork lift (2-5kw), and car (25-113kw). The “industry” is focusing on car sized fuel cells, and rightfully so, because everything in the transportation sector larger than a car can be powered by multiples of them, scale would be massive, and it is a major energy consumption sector. But it is the forklift size that is most appropriate for single family sized combined heat and power applications, and CHP justifies distribution pipelines and large scale dispenser manufacturing. It is pipelines that will drive down the end user price of hydrogen, and also permit large scale distributed solar. Small (“bicycle sized”and medium) fuel cells have a high potential use for emergency, camping, developing world energy. It also pairs well with paintball sized hydrogen connections for mobile power in robotics/battery centric applications or portable heat/welding. Something that has been too slow to occur is providing an affordable commercial fuel cell for integrators/application engineers to simply purchase.

As a precursor to pipelines, solar powered self supply for homes and businesses (and whole communities) is the cheapest source of energy and hydrogen. Storage through buried 4” FRP coils can support “curb side” dispensing taps that provide neighbourhood or apartment complex quick refuelling for vehicles and “paintball” tanks at “convenience pricing” which would allow homes and businesses to monetize energy surplus at 10c-20c/kwh, but also allow for easy access from pickup/delivery services to pickup large loads.