Steam Engine Update

Time for an update on the OSE Steam Engine project.

In the past month, I’ve pretty much finished up the technical aspects of what I am now calling the “Arrowhead Bump Valve” design, to distinguish it from the ball valve approach, which might be making a come back.  At Marcin’s urging, I encouraged several people to review the design.  Unfortunately, the current design suffers from several problems:

  • Horizontal layout of the engine will lead to uneven wear on the moving parts
  • Using high pressure steam at fast speeds is quite dangerous and includes the possibility of explosive failure
  • Operating at low pressure and speed seriously reduces the power output.
  • A fixed exhaust port in the cylinder means that air is compressed on the up-stroke, which is wasted work.
  • The arrowhead valve design may have trouble maintaining a seal and operate under full speed.
  • Bump valves will taking a beating and may fail.

While this is discouraging, it’s not the end of the project.  Still, Marcin has experssed some concerns:

Unless we have a proven design … I don’t feel comfortable on going forward, as that means years of development before product release. The key to success of our development method is its reliance on proven techniques. In their absence, we’re adding unnecessary risk.

Those who have been with the project a while know that the steam engine has already been years in development, so I can understand why he’d want to avoid more years.  Let’s take a closer look at some of those issues.

If you think about it, most internal combustion engines have a vertical orientation:  the crank shaft is often at the bottom of the engine (thought not always), pistons in a straight six are set up vertically (vee engines are off of vertical, I would note).  Back in the days of steam, engines were commonly laid out horizontally, but this started to change when people understood how the parts wear.


Since the piston is in fixed position with respect to the cylinder, if it moves in a horizontal plane, wear on the piston rings and any bearings will be on once side.  In a vertical orientation, the piston is essentially balanced with respect to gravity (assuming the engine base is level).  Wear is then distributed more evenly.  Furthermore, if the control valves are located at the top of engine, gravity works in our favor to keep the seal on the cylinder while steam is not being admitted.  Gravity also helps us to eliminate any water that may have condensed in the cylinder chamber.

I’ve gotten a healthy respect for high pressure steam (especially superheated steam – so called dry steam) as I’ve done my research.  This is a highly regulated technology governed by federal laws that determine how steam fittings are made, how boilers operate, what safety precautions are required, etc.  Escaping steam can cut through exposed flesh and leave very nasty burns.  It can also corrode and erode metal if care is not taken on the quality of water used.

Given the dangers of high pressure steam, I started turning my attention to lower pressures (under 100 psi).  Steam engines can work effectively with lower pressures, but put out less power (simple physics).  This leads to ganging cylinders for more output, but also increases the overall cost of the engine.  To some extent, it depends on what you want to use the engine for.  If you’ve got a small solar concentrator that puts out a steady supply of steam and you want to generate electricity for storage in OSE batteries (which are coming along, I note), then you might get by with a low power engine, perhaps geared up for higher generator speeds.  If, on the other hand, you want to generate a lot of power and use it directly – for an aluminum smelter, for example – a different engine design is likely called for, especially if you use correspondingly larger solar concentrators.

Valve design turns out to be it’s own sub-discipline in steam engineering.  In many ways, it is the heart of any steam engine design.  I’ve collected notes on valve designs.  The bump valve approach in the current design was strongly influenced by the White Cliffs Solar Power Station project conducted in Australia in the 1980’s.   The solar concentrator and steam engine developed for this project operated for years at high efficiency, low mainenance, and high power output.  Besides resident housing, it powered a nearby opal processing factory, similar in some ways to the manufacturing capabilities describe in OSE Global Village Construction Set.

Unfortuately, some of the key technical details were left out of the White Cliffs Report including what materials were used.  For these and other reasons, the design sifted from a ball shaped valve to an arrowhead shaped one.  We were also planning on making the engine from scratch, rather than converting an exising engine (a GM Lister Diesel in the White Cliffs case).

Recently, Marcin and I have been in communication with Mr. Ron Whaley, who was an engineer at Power Kinetics, Inc. and worked on the White Cliffs project.  Mr. Whaley has filled us in on some of the missing details left out of the report and has indicated a willingness to help us design a functional steam engine that will meet our needs.  Marcin plans to visit him at his home shop in the next couple of weeks.  This new information is likely to lead to some major design changes.  Until then, things are on hold.  We are not planning on prototyping the current steam engine design, as documented.

While Marcin explores new options with Mr. Whaley, I’ve been giving some thought to experimenting with steam engine design.  In software, when a new design concepts comes up, some one will often code up a prototype to see if it works.  It usually doesn’t take all that long – a few hours at most.  This is not the case with open source hardware.  The materials used in prototyping can be expensive and it can take a long time to fabricate (or buy) the parts needed to create one.  These are some of the reasons why we are holding off building the current design – too much time and too much money to spend on something that likely won’t work, in spite of what we might learn from doing it.

My thinking has led to the initial development of an experimental platform that’s designed to be less expensive, faster, and safer than building steam prototypes.  First off, I’m using compressed air instead of steam.  It’s easy to obtain from a commercial air compressor (I have one).  It’s simple to regulate the pressure (up to 200 psi) and it is much, much safer.  Secondly, I’m building a basic cylinder/piston system out of PCV pipe.  It turns out that the end caps for 3″ pipes fits almost exactly inside of a 4″ pipe.  It’s so close, I don’t think I’ll need piston rings or a seal of some kind – thought that remains to be seen.

I’m calling this an “air engine” and you can see more details of it at on the project page.  Marcin has indicated that the OSE process is not an experimental one, so I have moved this work to my own open source hardware site at the Phase 3 Project.  I’ve started blogging some of that work as well (see home page).  The first set of experiments with this platform will be the development and testing of a rotating valve concept.  Rotating valves are not a new idea, they were used in Corliss steam engines 110 years ago.  Rather that used a fixed timing mechanism, like an eccentric crank, my plan is to control the rotating valve using a stepper motor and a computer, such as an Arduino board.  If I can get a good seal and program it for rapid switching, I think we’ll be able to migrate this into the design of a truly Modern Steam Engine – one that uses a computer for timing.  This will enable, among other things, dynamic control over shut-off (the point when steam is shut off from entering the cylinder).  This degree of control allows for variable speed engines and simpler start up.  Adding a similar valve to the exhaust port and you can have dynamic engine braking.

This work is still in it’s very early stages.  If you are interested in supporting this work, let me know.  I have some very early experiments in mind before tackling the fabrication of the rotating valves.  There are also some intersting relationships to explore between the size of the steam flow channel and the diameter of the vavle.

Meanwhile, some good news.  My submission to the Open Source Hardware Summit to take place on Sept. 15, 2011 in New York City has been accepted.  While it won’t be a long presentation (ten minutes), it is an opportunity to make our work known to east cost hardware hackers.  The proposal reviewers seemed quite interested in the project:

The idea of an open source steam engine seems a bit inefficient at the surface but the logic behind it is actually quite intriguing especially when thought about it the solar heating context. I think this talk would a good shift in examples often cited in open source projects from the virtual to the real world. I would like to see more of the actual build process and the openness of the fabrication process rather than emphasis on the openness of the idea. Also the project has numerous CAD / specification type input while showing knowhow through videos / illustrations etc of the actual process would go a long way. Also the talk could touch up the idea of the Open Source ecology to make the steam engine part of it more relevant.

Open Source Ecology’s steam engine project would be an incredible addition to the Open Hardware Summit. Among many reasons, the project is very unique in the particular kind of ‘hardware’ it focuses on: a mechanical steam engine. Most submissions will probably focus on electronics, so the presenter’s focus on mechanics will undoubtedly stand out. If that weren’t enough, the mission of this project is of tremendous value to humanity. There are many important lessons that can be learned from the approach taken by the presenter and his colleagues at Open Source Ecology. The technology to end global hunger and poverty has existed for centuries, but this knowledge has not been effectively transferred to the developing world. Open Source Ecology has really nailed ‘appropriate technology'(E.F. Schumacher, et. al) and have seamlessly connected it to the open hardware movement. The Open Steam Engine appears to be a great representation of OSE’s work, and I highly recommend this presentation be included in the Open Hardware Summit.

Hopefully my presentation will be received in the same sense of enthusiasm.  My main message to this audience is that OSE has put a lot of effort into the development of a steam engine and though the current design has problems, we are continuing to work towards the creation of the worlds first open source steam engine.

– Mark Norton



  1. Robert MacLeod

    This is facinating. I look forward to learning more and following your progress. How do you plan to translate for underdeveloped countries and how will you get the interest for this equipment where it is most needed? What about Un Committees already workin in these areas. Wishing you all the best in September.

  2. Evan

    I look forward to hearing your presentation at OSHS next month. Your microprocessor controlled rotating valve is great! I myself spent many hours looking for solenoids suitable to serve as valve actuators but never found any fast enough or durable enough for the task. I didn’t think about rotating valves, there is real potential there.

    The old timers used some sharp engineering to build their engines, and it will be difficult for the OSE project to duplicate their successes. But we do enjoy some advantages over their efforts, namely: microprocessor control, superior materials, and the ability to machine exotic shapes. I believe, in the big picture, that in order to be successful, we’ll need to fully exploit these advantages in order to produce a simple to build, reasonably efficient and durable engine. It is also my suspicion that these advantages would most gainfully be employed against the steam turbine.

    Look forward to seeing you in NYC,

  3. dorkmo

    could you have the valves pull up and down, instead of twist? ive seen in ball valves in cases where they are 90% closed, that the area around the opening corrodes (from friction with stuff in the water i assume), expanding the opening to the point that it wont fully close. you kinda talked about variability, wasnt sure if you just meant timing, but just recommending against turning if you’re thinking about anything other than strictly full on/off.

  4. Joel

    I wonder if a DC motor/rotary encoder combination might be better than a stepper motor in this application. The valve itself will have some angular momentum, and under normal operating conditions, the motor will only be needed to overcome the friction of its bearings, and maintain a constant rate of rotation. It might even be worthwhile to have a flywheel on the valve, to make the system more stable.

  5. Sam T

    As both a historian of steam power and a believer in self-sufficient living, I find your project to be of immense interest.

    Have you considered an “oscillating cylinder” design? These eliminate the need for valves altogether. They’re used frequently in toy models because of their simplicity, but I think a few full-scale ones have been put to use as well. They are supposed to be much easier to build.

    Another option, if you’re worried about high-pressure steam, is to design an “atmospheric engine” that works by employing a separate condenser, as in both Newcomen’s and Watt’s designs. Admittedly, this would complicate your design, but in terms of safety you might be better off.

  6. DreamBuilder
    running off a compressor could generate electricity via alternator.
    So does the compressed air generator from
    they make the Air Car!!

  7. Lucas

    You mention that there are many design issues that you face. Now I’m no expert in steam (I just started reading up on this fascinating subject), but it seems to me that there is already alot of tried-and-true designs from the 19th century. These are effectively “open source”, simply because the patents have expired.

    Here’s some material you might find useful from the ever-stimulating Lindsay Publications.

    Essential Steam Power Library

    Steam Engine Principles and Practise

    The first three links (7 books) cover more the theory and aspects of steam engines. The last is a large volume of prints and engineering info.

    The books aren’t freely available, but I can personally attest that they contain very rare and valuable info, and are certainly worth the investment. Plus you’re supporting a great small-business.

    Check it out, and best of luck.

  8. Mark Warrington

    Stanley and Doble Steam cars had horizontal engines. I don’t see why you can’t go horizontal in yours.

  9. Mark Warrington

    Also, what about Lycoming aircraft engines and BMW motorcycle and old VW beetle engines that are horizontal???

  10. MontagK

    Hey Mark. Great update. Thanks for sharing. Following the project from Argentina. Heard of you guys a few months ago.

  11. Mark

    I’m happy that Marcin got in touch with someone who worked on the White Cliffs project. I think that Lister conversion done at White Cliffs was fantastic, and I suggest the basic design be adopted.

    High pressure steam in a small engine with a compact steam generator can be completely safe… it’s excessive superheat that would present the most potential danger. 500 psig steam with perhaps 100F superheat should be ideal (hell, the Stanleys were doing this more than 100 years ago).

    In my opinion, the goal should be the simplest possible configuration with a single cylinder uniflow using a single piston-operated bash/bump valve. No need for high power… in fact, I think one hp is enough as long as the thing can operate unattended for extended periods. The primary design goals should be low cost, simpicity, ease of repair, and reliability. High efficiency is desired, but not at the expense of the other goals…. and for God’s sake, do not incorporate any components that are not necessary (like arduinos or actuators). Adding unnecessary components to a system does not make it “modern”… it’s just plain silly.

    NOTE: I recently came upon a compact stainless steel ball check valve that might be used as a bash valve for a steam engine (just an idea). Consider that being able to thread the valve directly into the cylinder head of the engine will allow for quickly adjusting the position of the valve relative to the piston to compensate for wear and/or making quick adjustments during testing. Here is a link: (see the ball check valve on the bottom of the page)

  12. Tanel Välja

    Hello everybody.

    I recently joined the OSE project(even have my own profile page 🙂 ) and i’ve been checking up on the different types of work going on.

    While most of the stuff here is amazing i sometimes find projects that are trying to solve a simple problem the most difficult way possible(in my humble opinion).

    Considering that modern engines have taken almost a hundred years to develop(and the billions of euros of money put into them), there is little point in trying to convert a gasoline engine into a steam engine, on top of having fundamental issues(like effectively clearing back-pressure, so on the return cycle the engine does not have to waste energy to compress steam, etc) it is extremely difficult to manufacture, needs to have a huge amount of specific parts and is difficult to repair, should some catastrophic failure occur.

    Most of the technology here is geared towards easy reproduction and simplicity – why not oversimplify ?

    The Nikola Tesla turbine is a great example of a forgotten/lightly used technology that works, is extremely easy to produce, scales well and takes very little technical to assemble/create.

    It works with compressed air and steam(low pressure).

    easy to manufacture
    easy to test
    easy to scale(add either more turbines or enlarge the turbine blades)
    very safe

    need extremely good bearings(no joke)
    modeling air flows are difficult(to increase efficiency)

    More information can be found here:

    Enjoy 🙂

    1. Marcin

      Sounds good, but I think you forgot another con: ~30% documented mechanical efficiency as opposed to over 90% for simple steam engines. Please read our prior work on the topic on the blog and wiki.

  13. Tanel Välja

    There has been some amazing research going on during the past few years, with a few companies getting up to 40-50% efficiency.

    While carbon ceramic discs might be out of reach of the simpler folk(me included), i am planning to build 2 tesla turbines – but just like steam engines it will use 1 turbine with small blades(5 inches) with very high pressure and second turbine with lower pressure(8 inch blades).

    On top of just trying to replicate the results and improve efficiency by using “exhaust” steam, the goal is to also improve the blades themselves and how the blades are designed, mainly the exhausts.
    I’ve been doing fluidic testing and certain types of exhausts improve airflow between 2-3%. not much but over time it makes a difference :).

    I will try to compile my results by the end of october and then we will see if my idea of recycling the turbine exhaust will sink or swim…

  14. […] to know how far the car has evolved from being steam-powered coaches to modern practical vehicles.Driving a car back and fro from work has shown us our dependence on it. Read on to find the humble b…f Mercedes Benz) say that cars are light carriages for personal transport with three or four wheels, […]

  15. Mr S

    I can’t say I really understand the goals of this whole project you guys are doing.

    Is your plan that people in the developing world will replicate your machines and help them to build communities and beyond?

  16. evan

    A Tesla turbine is simple enough, but I could never find out how to calculate the boundary layer thickness of the steam against the disks. With out this data any design is just a really poor impulse turbine.

    Anyone know enough aerodynamics to apply it to steam?

  17. Early Car History Until 1840 | Car History

    […] cumbersome, and slow as well as expensive to maintain, and soon gave way to gas powered vehicles.The early automobile era played an important role in the development of modern cars. According to th…that Nicolas Cugnot, a French engineer, was responsible for the invention of the first […]