Recently, I published a guest post by Gene Preston which examined the electricity cost comparison for remote solar PV vs small nuclear reactors. This generated considerable discussion (128 comments), much of which focused on whether this was a useful comparison in many circumstances; what if, for various reasons, the small-scale nuclear battery is not a viable option?
Gene has since done further work to consider the problem of how to design a reliable, small-scale, low-carbon energy generation system, which is economically competitive (though not necessarily lowest cost). He uses a case study approach to consider five crucial aspects:
1. System 1: A rooftop solar and wind 100% renewables powered microgrid concept.
2. System 2: Like (1), but the 10 kW rooftop solar is replaced with 5 kW of centralized solar.
3. Analysis 1: Three ways to improve the reliability of a (nearly) 100% renewables system.
4. Analysis 2: The cost of CCS carbon capture and sequestration makes coal power uneconomical.
5. Analysis 3: Small nuclear power provides reliability without needing a new transmission grid.
First, here is a summary of the five cases. Following this overview, the case studies are given in full (for the more dedicated reader). I find these type of empirical studies incredibly useful in understanding the options available to us. Great work Gene.
In all the cases the microgrid has 150 homes. This number of houses was selected to best match the output of a 1.5 MW wind generator. Of course the size of the system could be scaled to any number of houses. The intent is to design each system to be as independent of the larger grid as possible. Each house has two PHEVs with 50 kWh batteries for a range of 100 miles of city driving for each fully charged vehicle. Each home is assumed to annually use about 12500 kWh plus another 12500 kWh for the PHEVs for a total of 25000 kWh per home annually. The PHEVs are assumed to be bi-directional power sources, being able to both receive power from the microgrid and deliver power to the microgrid in all cases. The microgrid consists of an undergound distribution system connecting the houses as well as the power sources local to the microgrid. All the costs for the distribution system, metering, etc that are the same for each of the above cases are not included in these calculations. The purpose of this analysis is to simply compare the cost and reliability of different types of power sources.
Here are the findings:
Case 1 has 10 kW of rooftop solar fixed panels at each house and a 1.5 MW wind generator for the whole neighborhood. The up front cost of the solar and wind per household is $90,000. This system will suffer occasional power deficiencies if operated as a standalone system. The interconnection costs for backup power from a larger grid were not estimated.
Case 2 replaces the rooftop solar panels with a centralized tracking solar system of size 750 kW.This saves each homeowner about $45,000 in up front costs and raises the question –- why are we installing rooftop solar when the centralized solar system is so much more cost effective? This system suffers the same problem as case 1 in that there will be occasional power deficiencies.
Case 3 looks at three ways to improve the reliability. #1 adds more battery storage and more renewable capacity to charge up those batteries and costs $100,000 more per household. However, this is still not a completely reliable system. #2 connects all microgrids in the US together with a massive investment in new transmission lines to gain reliability. The cost and environmental impacts are found to be impractical and the time to get approval and construct all the lines could take many years. #3 looks at installing backup generation at the microgrid instead of interconnecting. This is equivalent to firing up a put-put generator when solar andwind fail to produce enough power. The types of fuels discussed are oil, gas, coal, and nuclear. All of them are reliable, except they are deviations from our desire to be dependent only on 100% renewable power.
Case 4 looks at the cost of CCS carbon capture and sequestration and finds that it adds about 16 cents per kWh to the cost of coal generation, making coal unattractive as a base loaded source of power. Case 4 also shows that a 1 MW coal plant beside our subdivision eliminates the need forany solar or wind power at all and it would be the lowest cost if not for the CCS cost. With CCS coal looks no more economical than our 100% renewable plans, although the 100% coal is quitea bit more reliable than the 100% renewable plan, because the coal generator can run 24/7.
Case 5 looks at adding a small 300 kW nuclear plant beside the subdivision. It is air cooled and fits in a single homeowner lot. It silently runs for 30 years on a single fuel load and requires little maintenance. The wind generator is eliminated and the central solar is retained. Thesystem is reliable. The PHEV batteries are lightly used, allowing them to last longer. No new transmission lines are needed. This plan has a $45,000 up front cost to each homeowner.
Let’s consider a 100% renewables microgrid power system consisting of:
1) a single 1.5 MW wind generator located near a residential subdivision,
2) 10 kW of solar fixed rooftop panels on each home,
3) two PHEVs at each home with 50 kWh battery storage in each PHEV,
4) each home will use 25,000 kWh annually for electrical home use + transportation.
The wind generator costs about $2/watt or $3 million. The wind generator will be able to supply about 4000 MWh annually if the capacity factor is about 30%, which is typical. This microgrid might be able to operate independently from the larger grid if their location has enough wind and sunshine, such as Midland, Texas. Most customers will need to get their wind power from remote and windy locations that can produce energy at the lowest $/kWh cost.
The rooftop solar panels cost $7/watt or $70,000 per house and produce an amount of energy of (10 kW)(.77 DC-AC converter eff)(.15 annual capacity factor)(8760 hours/yr) = about 10,000kWh. The .77 is my EE friend’s new 4.4 kW system which produces 3.4 kW AC power.
The remainder of the energy must come from the wind generator, which is 15,000 kWh per home. In order to have some reserve, we should double the wind available energy as a part of the microgrid interconnection in which our renewables must also serve others so that we can draw power from other microgrids. Therefore, for estimating how many homes the 1.5 MW wind generator can serve, let’s be conservative and assume that each home will need 30,000kWh wind (double the 15,000). This means that our microgrid can serve a total of 4,000,000 kWh wind/30,000 kWh per home = 133 homes. Let’s round it off to 150 homes. The wind generator cost per home is therefore $3,000,000/150 = $20,000 which seems reasonable.
Each home will have two PHEVs in which most of the time one PHEV is active in driving locally and the other one remains parked in the garage most of the time. Each PHEV contains a50 kWh battery, which has a range of about 100 miles for city driving. Each PHEV charges at 220 or 240 VAC with a 10 kW load or 45 amps and can get a full charge in less than 5 hours andare charging when possible. These EVs are likely to cost about $40,000 each because of thelarge battery storage capacity and the battery cost of $10,000 for the electronics plus(0.4)(50,000) for the batteries = $30,000 total and then another $10,000 for the rest of the car.
The PHEVs are critical to storing energy for times when there is no wind and no solar, especially for an independent standalone microgrid system. The homeowner will need to be aware at all times of the charge state of the batteries and plan their daily activities around the power that is available from their batteries, their microgrid, and what’s available from the larger grid (if any).
The total homeowner cost of this system is $70k solar + $20k wind = $90,000 which should be affordable to most homeowners. The annual cost financed at 6% annual interest rate for 25 years is A = (90)(.06)(1.06^25)/(1.06^25-1) = $7040. The levelized energy cost is 704000 cents/yr /25000 kWh = 28 cents per kWh just for the wind and solar renewable power investment cost.
The 1500 MW wind generator is sized appropriately to simultaneously charge 150 PHEVs. Thesolar panel at each house is also sized appropriately to charge one PHEV at the house. A PHEVcould supply 1 kW power for up to 50 hours for backup power when there is no solar or wind.
Rooftop Solar Versus Centralized Utility Operated Solar
In the previous example, I had estimated that each home would need 10 kW of solar panels at an installed price of $7/watt and an annual capacity factor of 15%. The rooftop solar panels cost$70,000 per house and produces (10 kW)(.77 DC-AC converter eff)(.15 annual capacity factor)(8760 hours/yr) = about 10,000 kWh annually.
If we wanted to invest in utility-owned centralized solar and obtain the same amount of energy asour rooftop solar, how much would we need to spend?
The centralized solar cost is estimated to be $5/watt and have a 25% annual capacity factor. If 10 kW produces (10 kW)(.95 eff)(.25 annual CF)(8760 hr/yr) = 20000 kWh annually, we see that the centralized system produces twice as much energy as the roof top system. Therefore let usrequire only half the capacity or 5 kW per household to get 10,000 kWh annually for that home.
The cost per household is now (5000 watts)($5/watt) = $25000 versus $70000, which is a $45,000 savings per household. So why are we so interested in rooftop solar?
In the previous two case studies I used the batteries in PHEVs as the source of backup power when wind and solar power is not available, such as during a calm night. Windless nights will occur frequently. If we have too many windless nights and cloudy days in a row, our 150 homes willbe in trouble because the PHEV batteries will become run down and the lights will go out. And because the batteries are discharged, there will be no transportation either.
You might naively think that a simple connection to the larger grid will solve the problem. It won’t. I will discuss why below. Keep in mind Hawaii, which cannot connect to a larger grid.
#1 – The first possibility for improving the reliability of our 150 home microgrid would be to install more batteries. This will be an expensive addition because batteries are expensive. Doubling the size of the batteries in the PHEVs would cost another $60,000. To keep themcharged up will require increasing the size of solar and wind sources, possibly doubling them,which would cost each home owner another $20,000 for the second wind generator and $25,000 for doubling the size of the centralized solar farm (which is adjacent to the 150 home subdivision). We have spent an additional $100,000 to keep the lights on during extended calmand cloudy days. Our 150 home subdivision residents decide not to invest in additional solar andwind because the power supply is still not completely reliable, even with the additional battery,wind, and solar power additions. The additional storage idea is a bad idea.
#2 – The 150 homes may decide to connect to a larger system to provide backup power during the extended cloudy and calm days. However, the larger system is made up of thousands of microgrids just like ours, all hoping to draw on the larger grid for backup power, and hopefully not all at the same time. In this 100% renewables system, we have some microgrids that have extra power that can be used to supply energy to other microgrids that are short on energy. Each microgrid will need to install more wind and solar capacity than they need for their own system in order to have reserve power to assist their neighboring microgrid systems.
However, there is a severe shortcoming with this design of thousands of microgrids interconnected with each other. Because weather patterns cover large areas, we are likely to have times when large regions become deficient in power at the same time on cloudy calm days.This means that large transmission lines will be needed to cover the US, much like the interstatehighway system so that reserve power from one large area can be supplied to the other distant deficient area. These lines do not currently exist. They will be expensive and take many years to construct. There will be opposition to this plan due to its environmental impact and cost, so this plan may never be fully realized. Note that this interconnected system is not available toresidents in Hawaii. The building of all these lines connecting the eastern US to the western US to the Texas system (which are all currently isolated) is also a bad idea for improving reliability.
#3 – If the 150 homes microgrid wants a nearly 100% reliable source of backup power and does not want to connect to the larger grid, they could install a conventional generator that would onlybe run at times the renewables power is insufficient. There are four fuel types that could be usedto power the standby generators: a) fuel oil, b) natural gas, c) coal, and d) nuclear. Three emit CO2, except CO2 CCS (carbon capture and sequestration) might be used to capture the CO2. On Hawaii the backup fuel would probably be fuel oil rather than natural gas. The 150 homes might choose either a) or b) to keep initial costs low; however, these are not renewable sources.
In the previous example, #3c uses a coal generator to supply backup power to the 100% renewables microgrid system consisting of 150 homes. This would be a small generator of approximate size 150 times 5000 watts per house = 750 kW. Possibly a 1 MW sized coal plant would be a goodsize as a backup system. If the cost were $5/watt, then the cost of that backup system would be$25,000 per household. Because the capital cost of a coal plant is high, using it as a backup system does not make sense. That high a capital cost only makes sense if the coal plant were used as a base loaded generator. Interestingly, if the 1 MW coal plant were to run all the time, the solar and wind systems would not be needed and neither would the PHEV storage, except the battery storage can supply peaking power when the 1 MW generator cannot supply all the power demanded by the 150 homes, which would be rarely. Also, the PHEVs are going to be needed anyway to transition off the burning of oil and gasoline.
Ignoring the cost of coal fuel, the capital cost of the 1 MW base loaded coal would be a levelized annual cost of ($25000/home)(.06)(1.06^25)/(1.06^25-1) = $1956 per home. Then spreading that annual levelized cost over the energy consumed on average is 195600 cents/yr / 25000 kWh= 7.8 cents per kWh. Therefore, the base loaded coal plant supplying all the power is much lower in cost than the 100% wind-solar renewables system power cost, which was 28 cents/kWh.
But there is a problem with this design. The coal plant emits a lot of CO2. That CO2 will needto be captured and stuffed into the ground. Current estimates for CCS are about $100 per tonne (2204 lbs). A 1000 MW coal plant that is base loaded produces about 3 million lbs of CO2 per hour. However the CCS takes away 15% of the energy so that the 1000 MW coal plant is now 850 MW net electrical output. Considering that our coal plant is not 100% base loaded, but runs at an average power level of (150 homes)(25 MWh)/(8760 MWh) = 42.8% or 0.428 MW net electrical output, then our coal plant for the microgrid produces (0.428/850)(3,000,000) = 1511 lbs CO2 per hour on average or 0.6854 tonnes per hour.
The CCS cost is $68.54 per hour. On a cents per kWh basis the CCS adds 6854/(428 kWh) = 16 cents per kWh. Adding the CCS cost/kWh to the original coal plant investment cost/kWh we have coal costing 8+16 = 24 cents per kWh and that does not include the cost of coal fuel itself. Neither does it include the cost to pipe the CO2 to some remote injection point. The energy costof CO2 captured coal is nearly as expensive as our 100% renewables microgrid system. The only advantage of coal is that the power source is more reliable than the 100% renewables system, and that is why we were looking at coal in the first place.
Is there a better source of 24/7 power?
In this example, we instead use a nuclear generator to supply continuous power to the 100% renewables microgrid system consisting of 150 homes. This would be a small generator of approximate size 150 times 2000 watts per house = 300 kW that runs all the time. If the cost were $10/watt, then the cost of that backup system would be $20,000 per household. This provides a base load power source of sufficient energy to get past the cloudy calm days. Such a system would provide an annual energy of (300)(8760)/150 = 17000 kWh annually per home ormore than 50% of the annual energy needed. I will assume the nuclear generator actually provided 15,000 kWh annually to each homeowner. The wind generator could be eliminated from the mix of power sources saving the homeowner $20,000 in the cost of the wind turbine. The centralized solar farm could supply peaking power during the daytime and make up for the extra energy annually to get the annual 25,000 kWh annually.
The annual cost of the nuclear plant per homeowner would be (20000)(.06)(1.06^30)/(1.06^30-1)= $1453 annually and produce 15000 kWh. The energy cost is 145300/15000 = 9.7 cents per kWh. Nuclear also has an O&M cost that is about 1.6 cents/kWh bringing the total to about 11.3 cents per kWh for a small nuclear plant that costs 10,000 $/kW.
The cost of the centralized solar farm is $25,000 for 5000 watts per home, and produces 10,000 kWh annually. Its annual cost is (25,000)(.06)(1.06^25)/(1.06^25-1) = $1957 and the energy cost is 195700/10000 = 19.6 cents per kWh. Combining the solar and small nuclear plant costs produces an overall energy cost of (11.3)(15000)/(25000) + (19.6)(10000)/(25000) = 14.6 cents per kWh which is our lowest cost option yet. Note that we still have the PHEVs but the demand put on them to supply night time loads has been eliminated, thus extending the life of the batteries and saving a lot of money in transportation costs.
What about nuclear waste? The latest designs of small nuclear plants plan on using lower grade fuel and even burn what we would normally think of as nuclear waste as the main fuel of these plants. Therefore we create a new market for existing nuclear waste, and instead of throwing it away, we burn it further, getting much more energy out of the existing nuclear fuel, up to 100 times more energy. One example of a small nuclear plant is the Toshiba 4S plant.
There are several advantages of a small nuclear plant:
1) 24/7 reliable power nearly eliminates the need for transmission,
2) 24/7 hour base load operation makes wind power unnecessary,
3) the plant site is a small footprint,
4) the 4S plant is air cooled, not needing water,
5) the 4S plant is fueled once and runs for 30 years continuously,
6) solar and nuclear compliment each other in that nuclear provides base load and solardaytime peaking,
7) PHEVs have a continuous source of power by which to charge their batteries
8) the liquid sodium does not require a pressurized vessel,
9) there is enough fuel to power these reactors for hundreds of years using IFR technology.
10) once the fuel is spent, the entire reactor assembly is shipped back to the factory for refurbishing and another 30 year run,
11) the design is tamper proof eliminating the ability of terrorists to steal nuclear materials,
12) the design is operator error proof, i.e. the design is inherently meltdown proof.