By Brian K. Schimmoller, Managing Editor
One of the basic truths learned in college thermodynamics courses is that no heat-activated cycle can be devised that is more efficient than a Carnot cycle operating between the same temperature limits. In the 175-plus years since Sadi Carnot first described the externally reversible Carnot cycle, numerous scientists and engineers have labored to develop practical engines approaching its efficiency levels.
Remarkably, the Stirling cycle, whose theoretical efficiency can approach that of the Carnot cycle, was patented by the Reverend Robert Stirling in 1816, preceding Carnot’s writings by eight years. The Stirling cycle operates on a closed regenerative thermodynamic cycle, with cyclic compression and expansion of a working fluid at different temperature and pressure levels. In simple terms, the Stirling engine is a heat engine, relying on the heat produced in a combustion chamber external to the actual engine. As such, it can use essentially any fuel from natural gas to biomass and even coal. The heat is transferred to the engine’s working gas through the walls of the primary heater. The engine is a completely closed system. The working gas (which may be air or an inert gas such as helium or hydrogen) forces the pistons in the engine to move, compressing and expanding the working fluid, thus producing mechanical energy that can be used for motive work or to drive a generator and produce electricity.
Figure 1. The STM Double-Acting Stirling Engine. Photo courtesy of STM Power Inc.
Stirling engines have been developed in various configurations and sizes for numerous applications, ranging from hobby-size kit engines powered by the heat of the human hand to automobile engines and solar-powered generators. Commercial success has historically been limited by high cost, although Sears reportedly sold enough hot air Stirling engines fired by wood chips around the turn of the 20th century to warrant listings in their famous catalog for several years.
The past 100 years have been characterized by ample research but little market success. Engineering powerhouses such as Philips, Kockums, General Motors, Ford, McDonnell Douglas and MAN were all involved in Stirling engine development at one time or another, although most ultimately ceased research programs. A number of smaller, technology-focused companies have doggedly pursued development, including Stirling Thermal Motors (now STM Power), Sunpower, Tamin Enterprises, Stirling Technology Co., Whispertech, United Stirling and Stirling Energy Systems.
Skyrocketing demand for distributed generation technologies, coupled with the potential advantages offered by engines based on the Stirling cycle, have sparked renewed commercial interest in Stirling engines. While microturbines and fuel cells have garnered the majority of media interest with respect to the future of distributed generation, Stirling technology has a promising story to tell as well. Foremost among the benefits attributed to Stirling engines are its high efficiency, low emissions and fuel flexibility. The high efficiency is the result of the favorable thermodynamic cycle, while the low emissions and fuel flexibility are due to the Stirling engine’s external continuous combustion process.
“The majority of previous Stirling engine designs have been low wattage units,” explains John McKenna, CFO with STM Power Inc., which is currently pursuing alpha and beta testing of its 25 kW PowerUnit. “To succeed commercially, we decided we needed to take the Stirling concept and make it an industrial engine.” STM has made several technological advancements from previous designs to increase the economic and commercial viability of its product.
Earlier engine designs relied on two pistons in a single cylinder, the upper piston to displace the working gas and the lower piston to transmit the power to the crankshaft. This complicated the engine design because mechanical linkages were needed between the two pistons, and resulted in a low power to weight ratio, according to Lennart Johansson, President of STM Power. STM’s 4-120 engine (Figure 1) relies on a single piston per cylinder design. Each of the pistons in the four-cylinder engine is double-acting, providing both displacement and power. The upper portion of the piston receives the heat from the external combustion process, which increases the pressure of the working gas. By releasing the volume, the gas expands, moving the piston. As the gas expands, it is cooled in the lower portion of the piston, facilitating compression of the working gas and completing the cycle.
Older Stirling engine designs often depended on a rapid transfer of working gas into and out of the engine to accommodate load changes. This was inherently slow and further complicated the design. STM’s engine overcomes this limitation by using a variable swash plate, which converts the linear motion of the pistons into rotational motion. The angle of the swash plate can be varied on-line to accommodate load changes without changing the volume or pressure of the working gas. Idle to full-load operation can be achieved in one-third of a second.
STM has settled on hydrogen as its working gas because of its heat transfer capabilities, relatively low price and low flow resistance. At high temperatures, however, hydrogen is susceptible to permeation through the thin-walled tubes and to leakage through the mechanical seals. Hydrogen makeup, therefore, is necessary. To eliminate the need for an external supply of hydrogen, STM is developing a scaled-down electrolyzer (less than one inch in diameter) that will fit within the engine and generate hydrogen from water supplied to the engine. The electrolyzer is expected to consume less than 10 W of power, according to Johansson.
The STM 4-120 engine operates at heater head temperatures of 700-800 C (1290-1475 F), with a water/glycol cooling medium temperature of 50-70 C (120-160 F), resulting in a net electrical efficiency of almost 30 percent (LHV). Through the development and application of high-temperature alloys designed specifically for the engine, which will permit operating temperatures of up to 1,050 C (1920 F), STM’s Johansson expects to exceed a 40 percent electric efficiency level (LHV) in coming years.
The STM engine is particularly suited for combined heat and power applications. Unlike internal combustion engines, where about 33 percent of the input energy goes to electrical power, 33 percent goes to the exhaust gases, and 33 percent goes to the cooling water, the corresponding percentages for Stirling-technology engines are 35-40 percent, 10-15 percent, and 50-55 percent. For the STM engine, therefore, more than 150,000 Btu/hr (44 kWth) of 130 F hot water can be recovered using a water/water heat exchanger for use in cogeneration applications. Not all users will want or need hot water, of course, so STM and its partners are developing a radiator-equipped unit to dissipate the unused heat.
The external combustion system relies on a sophisticated burner design that incorporates recirculated, pre-heated air to completely burn the hydrocarbons and keep NOx emissions below 8 ppm (normalized to 15 percent oxygen) or 0.41 lb/MWh, which meets the 0.50 lb/MWh California Air Resource Board regulations and the 0.47 lb/MWh Texas Natural Resources Conservation Commission regulations for distributed generation products. Gaseous, liquid and dual-fuel burners will likely be available for the engine. “The dual fuel burner may be particularly valuable in the Northeast where dual fuel rates are common,” says Mark Fallek, vice president of marketing and sales with DTE Energy Technologies. “If users can switch to number 2 fuel oil in the winter months when the temperature goes below 20 F, they can get a nice break on their natural gas rates.”
STM’s decision to target the 25 kW product size derived from its involvement in a DOE research program to develop a solar dish/Stirling system with SAIC in the mid-1990s. The thermal output of the solar dish SAIC designed corresponded to a 25 kW engine. This size mirrors the size of most initial microturbine products. Recently, however, microturbine manufacturers have begun targeting slightly larger applications, in the 60 kW and up range, to capitalize on economies of scale and customer demand, and to minimize the effect of transaction costs on installed price and cost of electricity.
“I’m not sure there’s a market for any type of engine in the under 50 kW range,” says Dr. Steve Freedman, a Deerfield, Ill.-based power generation consultant formerly with the Gas Research Institute who has been tracking Stirling and other small-sized power generation technology for 40 years. “The transaction costs – permitting, siting, installation, and hook-up costs, both hardware and paperwork – may eat up the product’s value at the 25 kW size.”
The trick is to produce the engines at high-volume to capitalize on economies of scale. According to Freedman, “automotive quantities” of 30,000, 50,000, even 100,000 per year are needed to make the economics work. STM Power appears to be taking this concept to heart. “Technology is not our key focus right now,” says McKenna. “Cost is the main issue, and to attack cost, we are basing our manufacturing process on the lessons learned from 100 years of experience with the internal combustion engine.” The engines will ultimately be produced in assembly-line fashion using parts that can be sourced within a two-hour drive of STM’s Ann Arbor, Mich. offices, says Johansson.
STM’s hand-made alpha units are now entering the field for testing in a number of applications, including natural gas-fired standby power, black-start, waste heat, landfill gas, coal-bed methane, biomass, and hybrid bus. This will be followed in mid-2002 with about 50 beta I units, and another 300 beta II units in late 2002. STM Power has teamed with the engineering firm Ricardo plc to optimize the design and manufacturing of the 25 kW PowerUnits.
Stirling-technology engines face a two-pronged marketing challenge in gaining commercial acceptance. On the one hand, they must overcome the technical, commercial and general awareness head start enjoyed by other distributed generation technologies, particularly microturbines. The comparison shown in Table 1, along with the experience gained from the numerous planned field tests, is aimed at addressing this challenge. A telling distinction identified in Table 1 is that the STM engine’s output does not degrade with elevation. On the other hand, Stirling-technology engines must compete with reciprocating diesel engines in terms of manufacturing costs to achieve high-volume commercialization. Regardless of the market price charged for the STM engine, the bogey for the manufacturing cost is in the $400/kW range, says McKenna, comparable to that of reciprocating engines. As indicated by the recent demise of Honeywell Power Systems and its 75 kW Parallon microturbine, market acceptance is not a simple thing, reinforcing the criticality of low-cost, high-volume production.
DTE Energy Technologies is convinced of the STM engine’s commercial potential and has signed exclusive and non-exclusive agreements with STM Power to market the engine around the world. “I don’t think microturbines are that far ahead of Stirling engines in terms of market acceptance and penetration,” says DTE’s Fallek, “and when the Stirling engine’s higher efficiency and ability to run on low-pressure natural gas are considered, I think its long-term market prospects are better.” DTE will be offering STM’s engine to its distributors as part of its broad array of distributed generation technologies, which includes 75 and 100 kW internal combustion units, the 400 kW miniturbine it is developing with Pratt & Whitney Canada, and a 1 MW reciprocating engine. DTE intends to maintain a System Operating Center for its distributors and customers to remotely monitor, operate and/or coordinate maintenance for its DG fleet, dispatch to the grid, and sell power to the grid when pricing dictates.
The lack of long-term field operating experience is the biggest question mark surrounding the Stirling engine’s commercial acceptance. Both thermodynamic and non-thermodynamic factors need to be better understood. “While thermal cycling is hard to test for, it’s critically important to most end users,” says Freedman. “Moreover, extensive hours at high load and high efficiency are needed to demonstrate the engine’s long-term reliability, availability and maintainability.”
At least with respect to internal combustion engines, maintenance should be significantly reduced. The STM engine has about 50 percent fewer parts than a comparably sized IC engine since parts such as the inlet and exhaust valves, cam shaft, balancing shaft, ignition timing hardware and muffler are not necessary. Also, since the combustion byproducts are never in touch with the cylinders, and there is no metal-metal contact, the potential for wear is substantially reduced.
While maintenance is obviously dependent on duty cycle, STM Power expects its commercial engines will require only one preventive service check per year, after several thousands of hours of operation, and a major engine overhaul at about 50,000 hours. The preventive maintenance outage will entail tasks such as checking the air filter and ignitor, refilling the water if not connected to a continuous water supply for hydrogen makeup, changing the oil in the crankcase, and inspecting the filter for the hydrogen makeup system. The more extensive engine overhaul will involve tasks such as replacing seals, O-rings, bearings and the combustor.
Time will tell if the Stirling engine has what it takes to warrant investment in high-volume production capabilities to satisfy commercial demand. Its unique design and operational advantages, however, warrant consideration as another entrant into an increasingly crowded field of distributed generation technologies.
As interest builds in several states for establishing renewable portfolio standards, the developers of renewables-based Stirling technologies are hoping this is the impetus that drives their products toward widespread commercialization. Texas already has legislation in place mandating renewable energy generation levels, California is considering a bill that would require 20 percent of the electricity sold in the state to come from renewable resources by 2010, Nevada has passed a bill that would ramp renewable energy requirements biannually from 5 percent to 15 percent, and various other states are enacting or evaluating similar legislation.
While wind power is responsible for meeting a large fraction of the RPS requirements in states with such legislation, wind speeds are not sufficient in many states to generate enough power to meet required levels. In the sunny, arid southwest, for example, solar dish/Stirling engine systems may be able to fill the gap. These technically proven systems, which use an array of solar dishes to concentrate heat on a collector connected to a Stirling engine, can be designed for outputs from 1 MW to 1,000 MW.
Solar Dish/Stirling System. Photo courtesy of Stirling Energy Systems.
Stirling Energy Systems, Phoenix, Ariz., has been developing solar dish systems for more than 15 years and is convinced the technology is “ready for mass deployment as soon as sufficient investment makes mass production possible.” As with STM Power’s 25 kW PowerUnit, then, the key is high-volume production to reduce costs. Stirling Energy Systems estimates the cost to produce electricity from a 1,000 MW solar dish installation would be about 6 cents/kWh. Whether this cost is accurate, and whether similar costs could be achieved in smaller installations, is uncertain.