Stirling Energy & Commercial Power Generation – An Interim Step to Energy Independence

•February 28, 2009 • 4 Comments

For several years I have been studying and watching with great interest the development of the Stirling engine for use as a prime mover in stationary power generating applications and as a heat pump.  I was very excited in August 2005 about the press release announcing the ‘biggest solar deal’ in decades here in the Southern California desert that would utilize Stirling technology, and have been watching the developments in the North West with hermetically sealed free piston technology evolve from an idea to a finished product now just months away from commercial availability.

Stirling technology has been around for more than two centuries and although widely utilized upon its inception, has struggled to find a niche’ since the advent of the internal combustion engine (ICE). There have been successful implementations of Stirling technology and extensive research performed – especially back in the 1980’s – in programs and studies conducted by the Department of Energy, NASA Lewis Research Center, and Oakridge National Laboratories, to name but a few. (Many of these reports are available for download and make interesting studies.) A quick search of the Internet produces enthusiasts, bloggers and hobbyists who believe there is an application for this technology in something – somewhere – though much of what I have read might indicate that this technology as a resource may not be getting the attention that it deserves.

Today’s consciencious and educated society has created an environment rich with opportunities for technologies that are new, un-thought of, fresh and simply waiting for their time in history. With a government supportive of both the work ethic and spirit that made this country great and also supportive of the drive and need for us to be energy independent, it seems to me that Stirling is one of those technologies that may have a place in our near future.

There have been a handful of good books written about the “air engine” (“Principles and Applications of Stirling Engines” by West and “Air Engines” by Finkelstein and Organ being the must-owns for any student of Stirling technology) and only four or five companies around the world that have or are attempting commercialization of this technology on any scale. For as long as air engine technology has been around, there seems to exist a relatively small amount of available literature – even though there seems to be quite a lot of published scientific data! (Ask one of your friends at work if they know what a Stirling engine is!)

Having been a student of the Stirling cycle for a few years now, I believe it does have an application in our present time in history. Let’s consider a couple of the pros and cons. First a few pros; the engine is durable/rugged, few moving  parts, reliable, low maintenance cycle, can utilize low quality thermal energy (virtually any heat source), and is scalable in its design – though not linearly! Next, a few of the cons; the engine is not compact, high working pressure, not as controllable as the ICE, not commercially available – the biggest of all – it is still an engine!!

That last one is very important – it means that in order to get energy out, you still have to put energy in. When you truly consider the implications attached, unless we are discussing Solar Dish/Stirling, then Stirling energy is for the most part taken out of consideration as a renewable energy component! Or is it? Wow – that’s a mouth full!

O.K. – I’m going to leave it here for now but let me tell you briefly where I am going with this blog… In May, I will conclude a short period of investigatory research I am currently undertaking on the use of Stirling technology in commercial power generation.  I consider this research a prelude to other things to come in the future. There has been extensive research conducted by Oakridge into stationary applications already so we have a good idea of what the opportunities or capabilities might be. Through modeling I am considering the impacts at different locations to a traditional Rankine cycle (most coal fired steam plants), supplementary firing on the post-combustor side of the Brayton cycle and “stand alone” applications – what ever that might imply. I have completed some preliminary modeling of the Rankine cycle and the results are interesting.

Although I won’t publish my report or discuss my conclusions on the feasibility aspects of this study until May, I will blog some of my findings and thoughts  and certainly welcome any questions, thoughts, suggestions or just plain curious enquiries along the way…

Prelim. Conclusion

•March 18, 2010 • Leave a Comment

The Rankine and Brayton cycles were analyzed and nodes identified as candidates to insert a Stirling engine to capture waste heat and utilize it in a secondary process. In all cases, it was determined that the available thermal energy was of such a low grade that it was not feasible to apply Stirling technology to the system without some form of supplemental firing. Electrical output of a Stirling generator improves as the temperature difference between the hot and cold sides of the engine increases. 

Comparing additional thermal energy required and added to operate the Stirling system at its highest efficiency, the best candidate for an application of this technology is supplemental firing, post exhaust combustor, adapted to the Brayton cycle. By varying the mass flow of ignited fuel applied to direct heating of the pressurized exhaust stream between 0.05 and 0.13 kg/s, enough thermal energy is created to produce the temperature differences required for the modeled 25kW Stirling generator system used in our analysis to operate. 

Although Node 10, post-Intermediate Pressure Turbine, appears to be a potential candidate, the Rankine cycle is balanced such that any supplemental heating of the working fluid at this location may create additional thermal waste at the condenser downstream and negatively interferes with the Low Pressure Turbine as a generator. This creates issues on a larger scale with respect to the rejected heat. The Rankine cycle is not a good candidate for utilization of Stirling technology in a secondary application.

Almost twice the thermal energy is required to direct fire a Stirling generator as is required with supplemental heating in the Brayton cycle. This is significant when considering the commercial costs for natural gas would be double that of the Brayton cycle with supplemental firing as compared to direct firing. 

The Stirling generator analyzed in this application is producing 25kW of electrical energy. The thermal energy available as an input to a Stirling engine in aggregate would be enough to power one single 25kW system. This is an extremely small amount of generated electricity. To increase the amount of available electrical energy out, therefore increasing engine size, both air flow and mass flow of fuel in would need to be increased. Considering installed cost of the engine, and additional fuel for supplemental firing, the electrical energy generated is less than 1% of a 300MW Brayton turbine plant. Upon closer inspection, this is not feasible. 

In conclusion “waste” heat recovery from commercial power generating operations is not a viable application of Stirling technology. However, the Stirling system does deserve consideration for smaller scale, scalable or personal use applications. 

Initial cost, energy wasted, and the mitigation of the rejected thermal energy from the engine are three of the larger deterrents (local ordinances and code requirements not considered). 

The Stirling engine is unique in that it may use virtually any heat source, including unscrubbed biogas and solar. Stirling engines using natural gas as the external combustion thermal energy fuel source have emissions that are less than 20% that of an equivalent coal fired generating system. 

The tangible costs related to the Stirling system are in maintenance and initial systems capital costs, including costs related to procurement, fabrication and installation of the engine and all related engine ancillaries. It is generally agreed that Stirling engines will run for more hours with less maintenance required than conventional internal combustion engines. The cost installed for solar applications is $6.40/W for photo voltaic compared to $4.67/W for dish/Stirling (October, 2009). Currently, dish/Stirling is not suited for residential installation where photo voltaic are easily adapted to existing residential settings. Dish/Stirling systems are scalable but massive areas of undeveloped land are necessary to generate meaningful amounts of electrical power. The dish/Stirling system has only limited applications and viability.

The technology is not viable at this time for large scale supplemental commercial power generation projects. The technology could find uses at the individual or corporate user level.

more to come…

Last Round of Evaluation for this paper…

•May 24, 2009 • Leave a Comment

Well, I am at a point where I believe I have the data to support my conclusions.  I am planning to run one or two more comparisons. There have been some interesting results, not necessarily as I expected. Somewhat eye opening has been the overall thermal efficiency of the Rankine cycle and inability to capture more than ~36% of the available thermal energy. This is a cycle in which almost all the thermal energy is accounted for excluding low-grade, low pressure waste in the condenser. During my research, I have read a number of papers about Shinchi power station. Exemplifying Japan’s efforts to push power plant technology to its limits, with a measured thermal efficiency above 43% (LHV), Shinchi is one of the largest coal-fired power plants in the world using sliding-pressure technology, which avoids losses associated with throttling valves at low loads. High-pressure steam is maintained at 1,200 psig for loads up to 320 MW. As loads increase to 960 MW, high-pressure steam slides up to a maximum 3,700 psig. (Power Engineering, PEI, 2009) Expected to be commissioned through 2010 to produce in excess of 1,050MW.

As new plants come on line, the focus of my work continues to be, “what and how do we retrofit the existing ones – efficiently from both a cost and schedule perspectives “. It is an interesting problem. I believe there is an opportunity for Stirling technology and that it will play a role in our future.

It seems we return to the same challenges; the efficiency of converting the available energy in “fuel” through a process to electrical energy. Look out for a final summary of the results within the next three weeks. I will also make available a link to the entire paper once it is complete – perhaps I will also provide some of my earlier Stirling research through Cal State Northridge. 

Thanks for continuing to check in!

Additional Thermal Energy Required

•May 5, 2009 • Leave a Comment

 None of the candidate locations can provide enough thermal energy without supplemental firing of the thermal fluid. In the case of the Rankine cycle, it is the closed steam loop, for Brayton, the air/fuel mixture. 

In a ground based application, the Brayton cycle is defined by the appropriate entry Pressure and Temperature, with turbine stages added post combustor to extract every possible bit of energy from the working fluid. In this case, the working fluid is external to the engine, combustion is direct and air (containing oxygen) is added to fuel (a Methane mix) and the exit temperature of the combustor is the design parameter for the system. When the working fluid has passed through the final turbine stage, post heating must be added to the exhaust in order to give the exhaust flume enough thermal energy to rise up the stack and travel away from populated areas before dispersing. 

If a Stirling engine is inserted into the exhaust stream pre-combustor, there will not be enough energy in the exhaust fluid in order for the flume to rise. The only location the Stirling engine may be considered is at the secondary application point to utilize the remaining available thermal energy. 

The results of the comparisons for the three nodes in the Rankine cycle, supplemental firing in the Brayton cycle and direct firing are presented in this section below. A Compressed Natural Gas mixture with a HHV of 54,600BTU is used as a comparable fuel.

By considering the thermal energy available to each application and comparing against the required minimum temperature for the Stirling engine to operate efficiently, a calculation was performed to evaluate the amount of supplemental firing, and thus mass fuel flow, required.


The Rankine cycle, is a closed system, therefore thermal energy is added back into the working fluid by reheating the fluid passing it through a reheater or boiler. The Brayton cycle is an open loop and thermal energy is added downstream of the receiver directly into the exhaust stream. The latter principal is also the method employed when considering direct firing of the engine.

mheated air

mdotair = CpΔT                                                                                            mdot hot air = CpΔT

 By the 1st Law:

                               Q + mair Cp Tair inlet  –  mair Cp Tair exit  =  0

Q  =  (mfuel / mair) HVfuel


ΔT  =  Te  –  Ti  =  {(mf / ma ) HV} / Cpair

Nationally, consumer process for natural gas can vary between $0.55 and $1.23 per therm as determined by the supplier.


$’s / thousand cubic feet

Therms / thousand cubic feet

BTU / thousand cubic feet

ft3 / 10,000 BTU

$’s / 10,000 BTU






Source: EIA – Natural Gas Prices 2008

If a Stirling Engine with a working fluid of Helium and a hot side temperature of 540oC runs at 1800 RPM, the electrical energy output will be 25kW.

NODE   Tinitial     (K)   Tfinal        (K)   ΔT   mdot working fluid   (kg/s)   Δh air        (kJ/kg)   mdot fuel (kg/s)   Qdot     (kJ/kg)
8   594.9   813   218.1   372.2   743.0   5.1   276,561.3
12   610.3   813   202.7   60.5   690.6   0.8   41,780.0
14   318.8   813   494.2   311.7   1683.7   9.6   524,806.2
    498   813   315   58.4   316.1   0.3   18,460.4
    298   813   515   58.4   516.8   0.6   30,181.3


 The Solar Stirling or Dish Stirling System

The sun, with a heat rate of 440 Btu/ft2, and a twelve foot diameter parabolic mirrored dish focused to a receiver at approximately seven feet away from the center of the dish, provides the necessary 425oC to start the Stirling cycle. When discussing dish/Stirling technology, small variances in dimensions equate to large differences in the field. Over a distance of eight inches around the focal point, flux can decrease from 1200 W/m2 to 75 W/m2. The opposite is also true.

Photo voltaic is available to the consumer at approximately $6.40 / Watt; the proposed cost of a 3kW dish/Stirling system from INFINIA will be in the range of $13,500 – $15,400. (SOLO Kleinmotoren pricing of ~$40-50,000 U.S. for a 10kW dish/Stirling system.) At $14,000, this is equivalent to pricing available to the consumer at $4.67 / Watt. The dish/Stirling system can only operate 9 hours per day on average.


Continuing to Evaluate…

•March 14, 2009 • Leave a Comment


After modeling a typical, simple but accurate Rankine cycle steam plant producing 500MW of electricity, I chose four locations to consider as candidates to add a Stirling engine bank within the cycle itself. Keep in mind, the Rankine steam cycle is a delicately balanced system where most all of the energy is accounted for at every stage of the cycle. Not to be discouraged, I chose the following locations:


         the exit of the high pressure turbine (HPT)

         pre-intermediate pressure turbine (IPT) post reheat

         the exit of the low pressure turbine (LPT) pre-condenser

         a secondary leg, post reheat pre-IPT

         a secondary leg, adding reheat post LPT, pre-condenser


I modeled the nodes considering two scenarios:


  1. If no thermal energy is added back into the system, how are the overall output and cost parameters affected?
  2. If thermal energy is added back into the system, how are the same parameters affected?


It became immediately obvious that of the nodes I had chosen, the only two that may have any viability were the exit of the HPT and the exit of the LPT, pre-condenser.


The objective was to provide 1MW of power from the Stirling engines. The temperature of the fluid at the exit of the HPT was 322 oC. The temperature of the fluid at the outlet of the IPT was 45oC. I had built my Stirling model based on the geometry of the United Stirling 4-95 MkII, with a possible 25kW output operating at 20MPa with a working fluid of Helium.


At the HPT with no fuel added; the total system rating dropped to ~ 499,350kW, the plant efficiency decrease about 0.5%, the operating cost in fuel remained unchanged but there was a net revenue increase of about 1.2%. (This is the most interesting result.)


At the HPT with fuel added; the total system rating increased by 1MW, the plant efficiency decreased, the operating costs increased by ~ 0.5%, there was a net revenue decrease of ~ 0.75%  for an overall output increase of 0.2%. (Not viable.)


The LPT output is an interesting case. The temperature difference between the hot and cold sides of the Stirling engine would only be about 20 degrees, not enough of a delta T for the Stirling engine to operate. However, the working fluid at a very low pressure is still in a vapor state and must flow through the condenser prior to entering the low pressure pump. Lots of energy. What if we could replace the condenser, or reduce the amount of work the condenser does, by adding a large bank of Stirling engines to extract the energy from the working fluid? This would require further development of Stirling technology.


At such a low temperature, the number of engines required to produce 1MW would be excessive. This raises another question, what if we were to add a slight amount of reheat to the fluid prior to it entering the condenser? The results are devastating. If we increase the hot side temperature of the working fluid to 250oC, the plant efficiency decreases by >10%, the operating costs increase by >10% and there is was a net revenue decrease of >50% for an overall output increase of 0.2%. The condenser cooling water output temperature also increases 5 degrees. (Not viable.)


Initial conclusions: Adding engines without fuel and still seeing a net revenue increase seems to indicate the efficiencies of the Stirling engine bare further consideration. The increases are so small compared to overall plant losses, it does not seem viable. Adding engines and fuel did not produce a favorable result however, adding engines that operate on a very small delta T between hot and cold sides, at the condenser may warrant further investigation.


Modeling the Brayton Cycle. Thus far, adding a Stirling engine bank to the exhaust side of the cycle pre-stack has not produced favorable results. The gas temperature coming out of the turbine at 220oC is filled with carcinogens that contribute to acid rain if not disposed of in a combustion process. To do this, gas-fuel must be added to an afterburner. If just enough fuel is added to produce an adiabatic flame temperature of approximately 1450 oC (burning 100% of available oxygen), disposing of our ‘burnable gases’, and considering we must maintain at least 200 oC in the stack to fly the exhaust plume, there are only about 70K kJ/kg of energy left over at approximately 200 oC that may be fed into a secondary process.


What does this mean? It’s barely enough energy to produce 10kW of electrical output from a single Stirling engine using the same model as I have discussed above. At this point, the option is to add more gas-fuel. This further implies it may be equally efficient to consider the Stirling engine as a stand-alone generator.


Thus far, I am concluding that there is a very slight potential in the Rankine cycle pre-condenser, pending available engine technology, and there is little to no potential incorporating the engine in a Brayton cycle.


Let us further consider stand-alone and solar applications in the weeks to come.