Comment On Article About The Author More Articles By This Author

(Comparing electric power costs, Distributed Fuel Cell Generation vs. Central Stations) We want to compare the costs of distributed generation, specifically fuel cells, located at the site of the load or nearby, with the costs of electric energy from an integrated power supply system characterized by central stations located many miles away from the loads they serve, and connected to them by transmission lines, substations and distribution lines. A convenient place to start is to describe the central stations and their setting in the "grid". Tracing the history of the growth of the integrated electric power system in the United States will best illustrate the symbiosis of the central station and the transmission and distribution lines which combine to make up the “grid”. Central Station Power
For the last 100 years, electric power service has been supplied almost entirely by centrally located generating stations connected by wires to on-site loads. Prior to that time, if you wanted a supply of electric power for lighting or motors, you bought your own generator from the Thompson-Houston Co. or another vendor and located it on-site. Fuel cells may offer the prospect of returning to on-site generation or to locating the generation very near by the loads, in the center of a small collection of load sites or the "load-center". Contrary to popular belief, the first central station was not Edison's Pearl Street Station in Manhattan but was one owned and operated in San Francisco by The California Electric Power Co., one of the predecessors of Pacific Gas & Electric. It supplied electricity to arc lights in a near by area as early as 1879. The Edison central station on Pearl Street in New York was the first central station to serve customers using incandescent lights. It commenced operation in 1882. Thomas Edison was even better at public relations than at inventing so his central station is much better known. Both these stations operated using direct current or DC. DC has the disadvantage of being difficult and expensive to use if you want to change its voltage after the power has left the generator. Safety considerations dictated that the electric power should be supplied to the customer at a relatively low voltage, about 100 or 200 volts. Since you didn't want to change the voltage following generation, you would have to generate and distribute it at the same 100 or 200 volts. The low voltage promised safety, but when using DC, the disadvantage was that it could not be delivered very far away from the central station. The limit was about 1/2 mile from the central generating station. This followed from three principles. 1. If you want to deliver power and energy at a low voltage, you must deliver many amperes. For a given amount of power, the lower the voltage, the higher the number of amperes you must deliver. The converse is also true. The higher the voltage, the fewer amperes must be delivered. 2. The more amperes you wanted to deliver, the more copper wire was necessary to deliver them. The fewer amperes, the less copper you needed. Copper is expensive and you wanted to minimize its use. 3. To maintain a constant or near constant voltage at the system periphery so that your appliances would work whether they were near to or far from the generator, you wanted a voltage drop at the far end of no more than 5%. The further the distance from the generating station, the more copper was required to meet this parameter. The Transition to AC and the Growth of the Grid
George Westinghouse thought the answer to this dilemma was to use single phase alternating current or AC. When you used AC you could change the voltage easily by use of the transformer that had been invented by Gaulard and Gibbs. Westinghouse’s man Stanley built the first AC central station in Great Barrington, Massachusetts in 1886. Westinghouse wanted to emulate manufactured gas distribution systems. Those gas systems manufactured gas by passing steam over coal and compressed the manufactured gas to high pressures for delivery throughout a city, reducing it to lower pressures for safety to its customers before it entered their houses. A single phase AC system was built in Great Barrington, Massachusetts by Stanley, a former employee of George Westinghouse. It operated quite satisfactorily. It could distribute electric power through much of Great Barrington. It had one disadvantage. It could sell light and heat, but had great difficulty in starting electric motors with its single phase AC. Therefore it was a "light and heat" electric system but was not a "light, heat and power" system until Nikola Tesla invented the poly phase alternating current system. This started and ran electric motors very satisfactorily indeed. George Westinghouse bought his patents. After Tesla invented poly phase AC he developed ways of starting electric motors on single phase AC but the poly phase AC method of delivering electric power proved the best method and is still used for all electric motors except fractional horsepower motors as a general rule. Where poly phase, now usually three phase, service is not available, however, motors of up to about 10 horsepower can be utilized if the cost of extending three phase distribution lines to the site of the load is great. Alternating current service made it possible to deliver power at distances several miles away from the central station. The higher the voltage used for distribution, the bigger load center you could serve. The early distribution voltages were about 2400 volts or 2.4 kV. By the 1950s 13.2 or 13.8 kV was a typical distribution voltage. At this voltage a load center of about 5 miles in radius or 10 miles in diameter could be served from a small central station. In the last 20 years or so two or three utilities commenced distributing electric energy at 34 kV or 34,000 volts that formerly had been used as a "subtransmission" voltage. At this voltage, it was possible to distribute power up to about 25 miles away from the central station -- the load center had a diameter of 50 miles. Early on, the steam turbine entered the picture as the prime mover for the electric generator. The steam turbine-generator was powered by a steam boiler typically fueled by coal. The boiler and turbine were characterized by great economies of scale. In the 1960s, the Federal Power Commission estimated that when a steam turbine-generator unit size was doubled, the investment cost increased to only one and a half times the cost of the smaller sized machine, not double the cost of the smaller machine as you might expect. Larger generating units also burned less fuel per kilowatt hour. Commencing in the first decade of the twentieth century, this led to larger machines capable of serving first two and than many load centers. The 1920s were the heyday of building the "highlines", the transmission lines needed to integrate load over large areas so that the larger machines could be employed. Tesla's invention of poly phase AC again saved the day by permitting use first, of high voltage, e.g. 69 kV, 115 kV, 138 kV, 230 kV and then extra high voltage or e.h.v. of 345 kV, 500 kV and now 750 kV. Poly phase, almost universally three phase, transmission lines could gather the load from many load centers to make it possible to use these giant low cost, efficient generators. The small generating stations in each of the load centers were replaced by substations. They transformed the high transmission voltage down to the lower primary distribution voltage. As unit size went up, so did their fuel efficiency. From an estimated 8% efficiency in Edison's "Big Jumbo" machines at his Pearl Street Station in 1882, by the 1970s the fuel efficiency of generators 500,000 kW and 600,000 kW in size increased to 38% and up to 42% for supercritical units used mostly in Europe where fuel costs were greater. The cost of fuel is frequently the single greatest cost of generating power. Transmission voltage also went up to the extra high voltages. The Big Picture
These details are the trees in the electric power forest. Here is the key. What can be gleaned from looking at the forest is that it was much less expensive to generate from large central stations than from smaller ones. It was so much less expensive that even after the great cost of obtaining right of way and building all pole lines, mounting insulators and stringing conductors and building substations -- all needed to integrate the load, and even after incurring all the electric losses from those wires and substations, you would still come out ahead by using a large scale central generating station with its low per kW cost and its high efficiency. With the passage of time, what has happened to this forest? The ecology is suddenly changing. Now there are small scale generating units that are capable of generating at the same or even greater fuel efficiencies than large coal fired steam turbines. Efficient Generators Start Getting Smaller
The first one that came along in the mid 70s was the aeroderivative turbine generator. After the big increase in fuel prices in the early 70s, the airlines had pushed the turbine manufacturers to make their product use less fuel. The turbine manufacturers found they could use these same jet engines to power generators. The single cycle aeroderivative gas combustion turbines developed in the mid 1970s could operate at an efficiency of 42%. This product soon led to the combined cycle turbine generator taking the waste heat from the "front end" which was an aeroderivative gas combustion turbine, and collecting it in an HRSG or heat recovery steam generator and using it in a second cycle to generate even more electric energy with resulting efficiencies of first 50%, later 55% to 58% and just recently GE's "H" technology, claimed to operate at 60% efficiency. With gas costs only a little higher than coal, these aeroderivative gas combustion turbines and combined cycles were able to generate even more efficiently at sizes of 50,000 or 60,000 kW than the steam turbines ten times their size and therefore at a lower operating cost even though their fuel was somewhat higher in cost although GE’s H technology still required a minimum size of 400,000 kW. Because no boiler or coal handling equipment was needed they also enjoyed a much lower investment cost. Tiny Efficient Generators on the Brink
Now we find even smaller generators, fuel cell generators that can generate at the same efficiencies as the single cycle aeroderivative gas combustion turbine and at even a smaller size. A molten carbonate fuel cell only 250 kW to 300 kW in size can generate at an efficiency of 54%, a fuel cell of 3,000 kW in size can generate at 57%. Because these generators are sizes that match many loads or load centers, they can be used without transmission if they are located in load centers, and without either transmission or distribution if they are located at the site of the load. Moreover the fuel cell can also operate as part of a combined cycle generator. These are referred to as "hybrid" fuel cells that utilize their waste heat in gas turbines, with efficiencies that are revolutionary in electric generation. For illustrative purposes I will use the molten carbonate fuel cell of Fuel Cell Energy. Its products are currently field trial units in sizes of 250 kW, 1200 kW and 2400 kW but its commercial products will be sizes of 300 kW, 1500 kW and 3000 kW. It is also developing a hybrid fuel cell/gas turbine targeted at a product of 10 MW to 40MW. The electrical efficiency of its 300 kW commercial units is estimated to be 54%, 55% for its 1500 kW product and 57% for its 3000 kW product. Its estimated efficiency for a 10MW - 40MW hybrid generating unit is an amazing 78%. Currently its first costs are very high. Currently, its capacity costs for its single cycle 250 kW field trial units are about $5000 per kW, but estimated to be reduced to $4200 per kW for its commercial units even before it can engage in volume production. After engaging in volume production its costs are estimated to decline to about $1200 per kW. For its hybrid units the costs will be somewhat lower as the high fuel cell capacity costs per kW will be melded with the much lower gas turbine costs. In contrast for the central station, coal fired steam turbines and gas combustion turbines are far lower in first cost. Coal fired steam turbines, including their boilers and coal handling equipment, average a base cost of about $1000 per kW in the US, plus however much one has to invest to satisfy air pollution control requirements; that may be a lot. Gas combustion turbines or combined cycles are a lot lower in cost. They may cost only $350 to $500 per kW. But their output goes down considerably on a hot day so that would make their costs a little higher. Also, the advertised efficiencies of the aeroderivative gas turbine and combined cycle are those for efficiency at full output. I am advised that at partial output their efficiency declines significantly. Average efficiency for a single cycle aeroderivative gas combustion turbine generator, over the annual load curve may be as little as 32% according to one author. However, in comparing the first cost of central stations in the grid with comparable first costs of distributed generation located at load sites or load centers, one must add to the cost of the central station the "wires cost", the cost of the additional transmission, subtransmission, substations and primary and secondary distribution needed to deliver the power from the central station to the customer's meter. According to a study by Detroit Edison, within the last few years, its "wires cost" has rapidly increased from a historic average cost of $500 per kW to its current long run incremental cost of about $850 per kW. A nationwide study carried out by A.D. Little, shows even more disturbing news. In that study the historic average cost is also $500 per kW but the nationwide average for new construction, a broad average which could encompass Detroit Edison's results previously referred to, is $1,290 per kW without substations, and one must add from $50 to $300 per kW for substations. Those are investment "wires costs". The operating costs are electrical losses of some 13% to 16% electrical losses during utility on-peak periods and of course operations and maintenance costs. Load Diversity
With this background we are almost ready to evaluate the comparative economics of DG vs. the Grid. But first we must discuss the subject of load diversity. If I operate a simple hypothetical utility with only two customers, and each of my customers takes 100 kW of service all day long, I must invest in, maintain and operate 200 kW of capacity with necessary reserves for forced outage. But if Customer A says he only wants power during the AM hours, and Customer B tell me she only wants power during the PM hours, I can get by with only 100 kW plus reserves. We call this the benefit of load diversity. Alas, life is not ordered so conveniently. A utility has many customers. All will want service on call 24 hours a day but, with the exception of industrials operating two and three shifts, and commercial customers open late into the evening, during the late night and early morning hours none of its other customers will actually use very much electric energy. Their needs will increase during daylight hours but will likely peak at different hours during the day and early evening. For low load factor loads such a residential loads, the load diversity in a collection of diverse residential loads will reduce my capacity costs greatly. Each residential load may typically peak at ten kW. However to serve the coincident peak, a utility may need only one kW of generating capacity to serve each of his many residential customers whose loads are integrated by transmission and primary distribution lines. Adjusting Central Station Efficiency for Electrical Losses In Transmission and Distribution.
Utilities loads vary hourly and daily and encounter their highest loads during distinct “on-peak” periods during weekdays of about 10 hours to 14 hours in duration. A weekday on-peak period of 6:00 AM to 8:00 PM would not be unusual. Most utilities in the United States find these daily on-peak periods highest in the winter, and since the onset of air conditioning, in the South many have summer peaks. Some have rotating summer and winter annual peak hours with each annual winter peak being followed by a later even higher summer peak hour and that peak followed by a higher winter peak. When the transmission and distribution systems are fully loaded, their electrical losses are the highest. One utility, Detroit Edison, has stated electrical losses in transmission and distribution reduce energy delivered to the customer’s meter from central stations on average during on-peak periods some 13% to 16%. That would mean that if efficiency is measured at the customer’s meter, it would be only 87% to 84% as great at it was when measured at the low voltage bus of the output transformer at the generating plant site. So a central station combined cycle generator with a full load efficiency of 50% when measured at the generating plant, would have only a 43.5% or 42% efficiency when measured at the customer’s meter. A steam turbine with an average efficiency of 33% could have an efficiency of only 27.7% at the customer’s meter – a modern 500,000 kW central station steam turbine unit with an efficiency of 38% at the plant site at full load would have only a 32% efficiency when delivered from a distribution line many miles away. On average, coal fired steam turbines have an efficiency of 33%. At the meter this turns out to be 27% to 28%. Even GE’s H Technology gas combustion turbine combined cycle with an efficiency of 60% at full load will encounter these T & D losses because it will have a minimum capacity of 400,000 kW and will need transmission and distribution to distribute its output via transmission to much smaller load centers and from those distribution stations by distribution lines to even smaller on-site loads. Its vaunted 60% combined cycle efficiency at full load, with 16% electrical losses subtracted, turns out to be 50 delivered to the load. In contrast a load center of only 10,000 to 40,000 kW could enjoy an estimated 78% efficiency with an on site hybrid molten carbonate fuel cell. Of these losses, most take place in the lower voltage distribution lines. Transmission losses may measure only 3% to 5%; low voltage distribution losses add the rest. Losses are much lower when averaged with off peak periods over a 24 hour seven day week. Losses on foreign systems may be greater than those in the United States. Each year, according to WADE, the World Alliance for Decentralized Energy, energy losses from the world's transmission and distribution (T&D) systems exceed the combined annual electricity consumption in Germany, the UK, Spain and France. The rate of loss (measured on a 24/7 basis I think, where on-peak and off-peak losses are averaged) is 7.4% in OECD countries (a poor performance) and 13.4% (an appalling performance) in developing countries. According to WADE, losses in some countries are especially frightening. Nigeria - 33%; Colombia - 24%; Haiti - 57%; India - 23%; Albania - 54%; Kyrgyzstan - 32%. These losses - directly associated with the traditional model of central power generation – are, according to WADE, causing massive social, economic and environmental damage to the world's poor countries. It claims that “Millions of people are failing to receive a supply of electricity as a consequence; national fuel bills are billions of dollars higher than they could be, and pollutant emissions are causing untold health and environmental harm.” On-site distributed generation, including fuel cells have no losses, zero T & D losses. If the generation is placed at substations, the low voltage distribution losses can still be as much as 9% to 11% measured on-peak; much lower on a 24/7 basis. Comparing On-site Fuel Cell Power with Grid Power
Now we have most of the tools we need for the comparison we want to make. We can compare two distributed generation customers with the same customers who are served by the grid. One DG customer, Customer A, serves his apartment house or office building, or college or hospital or brewery, with a high temperature 300 kW molten carbonate fuel cell or solid oxide fuel cell. Another, Customer B, connects his single family residence served by a small PEM fuel cell with relatively short low voltage distribution to a microgrid, a tiny grid of primary distribution lines for the benefit of load diversity and reserve sharing. Or, he may connect to a larger scale molten carbonate or solid oxide fuel cell installed to serve his subdivision, or installed by a utility at a nearby substation. The other two, Customers C, owns a large apartment house or otherwise needs service in a range where his load is usually at 300 kW or more, and Customer D who owns a single family residence, choose to retain their service from central stations through the grid of transmission lines and via substations and primary and secondary distribution facilities. Let us start with the apartment house owner interested in fuel cells, Customer A. At the present time, his initial cost of an MCFC fuel cell will be about $5,000 per kW for a field trial unit with a capacity of 250 kW and $4,200 for a commercial unit with a slightly larger capacity of 300 kW. With volume production, the manufacturer claims these costs are expected to decline first to $2,800 per kW at 50 MW of annual production, and then to $1,200 per kW with 400 MW of production. The sum of $1,500 has also been mentioned as a possible first cost. If compared with the first costs a utility might incur for a generator for Customer C, these will look high. The central station coal fired steam boiler and turbine will cost only $1,000 per kW with more depending on how much pollution control equipment is desired. Its efficiency, for an optimally sized 500,000 kW or 600,000 kW coal fired steam turbine will be only 38% If the cost of natural gas is only a little higher than coal per million BTUs, a single cycle gas aeroderivative combustion turbine might be even more desirable at only $350 or $500 per kW because no boiler is required, only a turbine and generator. But these are first costs per kW of their output at the generating plant. Because customers loads are integrated by transmission and distribution and because there is diversity in their loads, the coincident peak of 100 apartment house customers, C will be far lower, perhaps as low as only 30% of the total non-coincident peak. So instead of $1,000 per kW to satisfy the sum of non-coincident peak load requirements, we need only 30% of that or $300 per kW of the sum of non-coincident peaks. However to obtain the advantage of using that large scale unit, we must install transmission, distribution and substations. On a broad national average these will cost about $1,300 per kW for transmission and distribution, and some $50 to $300 per kW for substations. And this cost is rapidly rising. To serve the customer with a molten carbonate fuel cell, his cost will now be $4,200 to $5,000. When production volumes are 50 MW per year his first cost is anticipated to decline to $2,800 per kW and at the bottom of the production cost curve with a production volume of 400 MW per year, his cost is expected to be only $1,200 per kW. To recapitulate, Customer A's initial cost for distributed base load generation is $5,000 per kW for field trials, and $4,200, $2,800 or $1,200 per kW for commercial units depending on how far into mass production his supplier is able to get, and his fuel efficiency is very good. If natural gas rises precipitously, relative to coal, he can run his generator on the output of a coal gasifier. His generator is very near his load so he can also benefit from its heat energy for domestic hot water or space heating or air conditioning without any addition energy cost. His combined heat and power CHP efficiency may be as high as 80%. Customer C's generator cost is 30% of $1,000 (plus added cost for air pollution control) which is $300+ and only $105 to $150 for generator costs for the single cycle combustion turbine or combined cycle. But to those costs must be added some $1,500 for transmission and distribution costs, bring the total to $1,800 for coal or to $1605 to $1,650 per kW for gas combustion turbines. We must raise Customer C's generation costs a little higher because of electrical losses of 13% to 16% during utility on peak times. Therefore for each kW used by the customer, the utility must use 1/ 0.84 kW or about 19% more generator capacity because of the power loss at the customer's meter. These will bring the central station generator's cost up slightly from $105 per kW to $125 per kW for the lowest cost gas combustion turbine and from $300 per kW to $357 per kW for the coal fired steam turbine without pollution control. To these we add $1500 per kW in transmission and distribution costs to get a total of $1605 to $1857 for Customer C, taking his power from the grid. The central station is too far from the load for it to use its heat. The central station heat must be exhausted into the air. Compare this with the $2,800 or $1,200 in first costs from an MCFC fuel cell which Customer A will encounter with volume production of fuel cells. TIAX, the successor to AD Little, predicts costs for small planar SOFC modules assembled into 250 kW assemblies with an electrical efficiency of 50%, can be manufactured for as little as $450 per kW with large annual production volumes, i.e. 2500 MW per year. Installed on site, they will incur no first costs for transmission and distribution and no costs for 13% to 15% energy losses during utility on peak periods. Because his fuel cell is located in his cellar or back yard, the fuel cell customer can use the heat generated as a byproduct of his electrical generation for domestic hot water, space heating or air conditioning without any additional fuel cost. With the right kind of interconnection contract, he can dispatch his fuel cell to provide heat and if the electrical generation is greater than needed, can sell it to his local utility. Customers B and D are the residential customers. Customer B will buy a small PEM fuel cell. He can use its thermal output for domestic hot water or space heating. Doing that will bring the CHP or combined heat and power fuel efficiency of the device up to 70% or 80% from only 29% as its electrical efficiency. His cost for a fuel cell may be as low as $400 per kW at the bottom of his production cost curve or a first cost of $4,000 for a 10 kW load. It is much higher now. Customer D stays with the grid. Residential customers will likely have greater load diversity than apartment house owners. To serve the coincident peak, only 10% of the generator capacity will be required. So if the utility installs a coal fired steam turbine, its generator costs per kW of the sum of non-coincident peaks will be $1000+ (pollution control adder) or $350 to $500 per kW for a single cycle aeroderivative gas combustion turbine or a combined cycle unit. Adjusting for load diversity will bring this down to $100 per kW of the sum of non-coincident peak for the coal fired steam turbine or only $35 to $50 per kW of the coincident peak for aeroderivatives gas turbines. Adjusting for on peak average losses, will bring these numbers up to $119 for the coal fired steam turbine (no pollution control) and $42 to $60 for the aeroderivative gas turbine generation capacity. This brings Customer C's utility costs which it must pass on to its customers of $1619 ($16,190 for a 10 kW residence) when it passes on the costs of a generation and transmission expansion program with coal fired steam turbines, or $1542 to $1560 (ten times that for a typical household) if it expands its system with natural gas fueled combustion turbines. If he wants space heating, domestic hot water or air conditioning he must buy more energy from his electric utility or gas from his gas utility. It should be noted that most of Customer D's first cost is transmission and distribution. Very little is for generator capacity. Adding to Distribution Substation Capacity in Small Increments. More than half of the investment in transmission and distribution is made at the distribution level. However installing fuel cells at a utility's distribution substation is very attractive, even though the savings in wire cost will not be as great. According to EPRI, there are strong incentives for installing fuel cells at substations also. The foregoing examples showed fuel cells being installed at the load sites or in a subdivision very near the load. Fuel cells might also be installed by a utility at its distribution substation that has grown to the limit of its power supply. If it is limited by the transmission lines input to the station, the utility will have to make expensive expansions which will likely be made to take care of needs for the next five or ten years. If only 300 kW is needed for the first year, the utility will have a lot of expensive investment for which it is incurring costs that will not be earning revenue for a long time. The same would be true for a utility that has added cooling fins and fans to its transformers in a distribution substation and now finds it must replace the transformers with larger ones. It would be impractical to add transformers which would only increase the capacity by a small amount. The utility would want to add capacity sufficient for the next 5 or 10 years at the least. So until the end of that period, the investment may not earn a full return. According to EPRI, this might be an excellent niche market into which larger scale fuel cells could fit, even before their decline in price. Conclusions
When you are looking at a price entry point for entering a market with fuel cells for customers to use on site, look principally at the cost of the transmission and distribution necessary for load integration so that central stations can be used; don't use the cost of the central station generators as that market entry point. In some instances, fuel cost savings with fuel cells operated as base load generation can overcome the disability of high first costs. This is more likely on islands, in steamships, and in remote small load centers that cannot obtain scale economies using large central stations because they cannot integrate their load center with other load centers. Utilities may find larger scale fuel cells attractive at specific distribution substations, particularly in low load growth areas, even though their first cost is greater than the utility's average cost of a generation and transmission expansion program using conventional resources. When the generator is tiny so it can be located at or near the load, its thermal energy can be put to use for domestic hot water, space heating and air conditioning -- so count those CHP savings for the fuel cell.