Ambitious ORC Turbine

[Toymaker]

For the past year I've been slowly designing and building pieces of a rather ambitious "steam" engine project. I'm retired so I have the luxury of taking as much time as I want, or need, to complete this project, and like most folks here, this is my hobby, not my job.

My goal is to build the smallest, lightest turbine possible that develops around 200 HP (150 kW) with the best fuel efficiency that I can achieve, and the entire system must also be mobile.

With those requirements in mind, I set out to build an ORC (Organic Rankine Cycle) turbine engine. I've already posted a few pics of some of the parts in the General Engine Discussion forum and the Boilers forum, but I'll start from the beginning in this forum as this project is still incomplete.

I started by making a list of all the various parts I would need, then sketched out a system diagram and finally drew up a CAD drawing which made making changes easier. Keep in mind that the below drawing is still in flux and incomplete, but I believe it represents at least 95% of the finished system.

For several reasons I decided to use a 3 stage axial flow turbine. This is a closed system which constantly re-cycles the working fluid, which in this case is not water, but an organic liquid instead. Below is a rough sketch of the turbine and other parts, to be discussed later. I've left out a lot of detail, which I will discuss if asked.

Turbines can be tricky to regulate, especially small ones. They can over-rev in the blink of an eye. Also, I've chosen to use a supercritical boiler with high flow and heated with a forced air burner with high flow rate. For safety and for best control, I decided to use a microcontroller along with sensors and servo motors to take full control of the entire system. These on-board computers are often referred to as a FADEC, for Full Authority Digital Electronic Control. Below is the FADEC system I will be using.

This was my starting point and continues to be my guide for what still needs to be accomplished.
Again, please keep in mind that these drawings are still in a state of flux, and can be changed and added to as needed.

Next, I'll post a few pics and videos of the actual parts I've managed to make thus far.

 

[Toymaker]

One of the advantages of using an organic working fluid instead of water is that the turbine can be made with only 1 to 3 stages and still be quite efficient. I chose to build a 3 stage axial flow turbine with most parts made from 6061 aluminum, including the blades. The center shaft, or spindle, is steel and both bearings are ceramic; (The bearing shown in the pick below is standard Stainless Steel, but it has already been replaced with a full ceramic thrust bearing).

The pic below shows the outside of the turbine housing. Those 8 large threaded holes drilled at a 45 degree angle are where the "steam" flows into the steam chest and from there through the nozzle vanes. The 4 socket head screws penetrate through the housing and hold the nozzle in place.

The 3rd and largest Blisk (Bladed Disk) is just over 3.5" in diameter while the 1st stage blisk is 2.36" in diameter. All three blisks slide onto the splined spindle. Below are the 2nd and 3rd row blisks and the second set of stator blades.

The pic below shows how the boiler output "steam" is routed through eight 8mm tubes into the turbine housing.
Eight additional 8mm aluminum tubes will be brazed into the holes you see in the uppermost oddly shaped block and then wound around in a spiral pattern forming the super critical section of the boiler.

 

[Toymaker]

The burner uses as cordless leaf blower to force air through the combustion chamber. The design is similar to the burner section of a jet engine, and the burners inside of "torpedo space heaters" and burns standard diesel fuel.
The fuel nozzle is a siphon type that uses low pressure air to suck in the fuel and atomize it into a fine mist; at 6 psi air pressure, the nozzle is rated at 14 liters/hour. My manual testing has shown the burner to work well up to 7psi. The video below shows the burner running at near Maximum fuel burn. That no black smoke is seen from the exhaust indicates 100% clean fuel burn, or at least very close to 100%. 14 liters/h = 148 kWh = 200 HP-hours.

 

[Toymaker]

The burner is made from two different sizes of stainless steel cups which are widely available here in Thailand. The handles are removed with my shop grinder and bottoms are cut out on my lathe. One cup has swirl vanes cut into the bottom along with holes to mount the fuel nozzle.


Two rings of swirl vanes, like the one shown below, are added to help force the fresh air to swirl around the outer surface of the inner burner cans to keep them cool and help guide the fresh air into the combustion chamber in a swirling pattern. The swirling tornado-like flames stay inside the combustion chamber just a bit longer, allowing for greater fuel-air mixing and more complete combustion of all the fuel. The last pick in this row shows all three outer shell cans stacked together. The silicone tubes carry low pressure air and siphoned fuel into the nozzle.




This last pic shows how the inner burner cans fit inside the outer shell cans. The two sets of swirl vanes keep the inner cans in alignment as they expand and contract during heating and cooling.

Now before someone asks, "why didn't I just use a couple of stainless steel tubes or even rolled stainless sheet",...the answer is that I've had to learn to use what's available while living in Thailand. Stainless tubes and sheet are both very hard to acquire where I live, so I use what I can get. But as an un-planned benefit, turns out that the separate cans make disassembly and re-assembly very easy.

Finally, here's a pic showing the small cordless leaf blower mounted onto the burner where it blows fresh air into the combustion area. The battery is removed from the blower and it's electric motor is powered by a PWM motor speed control which allows me to manually increase or decrease air flow into the burner.

This last pic also shows the two white porcelain electrodes for the high voltage spark ignitor.

 

[Toymaker]

In preparation for using a FADEC to control the engine, I modified the manual pressure regulator I'm using to control fuel flow, by adding a stepper motor and electronics to drive the motor. The manual regulator is on the left while the FADEC controlled version is on the right. Same regulator and gauge in both pics, just added some automation goodies to the right pic.

 

[ddmckee54]

This looks too well thought out to be an impulse build. What are you planning on doing with this thing? Nice looking machine work, especially on the blisks. What did you use to cut the internal splines?

 

[Toymaker]

This looks too well thought out to be an impulse build. What are you planning on doing with this thing? Nice looking machine work, especially on the blisks. What did you use to cut the internal splines?

Thanks for the compliments on the machine work,...it's all done on my DIY CNC Lathe and Mill. I think it's been about 8 or 9 years ago that I built my CNC Milling machine, and with that and a CNC rotor tool I also built, I'm able to machine most everything I need, including those internal splines. Here's a link to a YouTube video I made months ago on how I did it: Cutting Internal Splines.

You're right about this project not being an impulse build, the engine has a very specific purpose. Most folks look at the hot flames coming out of a jet engine and believe it's all those hot gases that make the thrust, and that is true, but a jet engine's exhaust doesn't need to be hot in order to produce thrust, the gases just need to be moving and have mass. F=ma governs how much static thrust a jet engine produces; the mass of the exhaust gases times it's acceleration is how thrust is calculated,....and it's the compressor inside the engine that is doing all the work to produce that thrust. Essentially, jet engines are just big air compressors that direct their compressed air out of the back of the engine.

It really doesn't matter what type of motor drives a jet engine's compressor, as long as that motor has adequate power and can spin the compressor at the required rpm the compressor will produce thrust.

So that's my first step; determine how much thrust a steam turbine can produce. If all goes as I hope it will, that's the first step in an even larger project.

 

[ddmckee54]

What kind of rpm's are you aiming/hoping for from the turbine? Have you done any dynamic balancing on it? (Or maybe just haven't gotten that far yet?)

 

[Toymaker]

What kind of rpm's are you aiming/hoping for from the turbine? Have you done any dynamic balancing on it? (Or maybe just haven't gotten that far yet?)

Planned max RPM is 70,000.

Currently, static balancing only. I did find an interesting technique for dynamic balancing using a standard smart phone: I haven't used Dynamic Balance App but it was written to help drone owners balance the motors on their drones but I should be able to use it tell me if I need to balance my rotor or if it's close enough I needn't worry.

If you know of any good methods for dynamic balancing, please do let me know.

Two methods of static balancing that have worked very well for me in the past: One method requires both shaft ends of the rotor to be placed on knife edges which must be perfectly horizontal. The heavy part of the rotor will slowly roll to the bottom. The second method replaces the knife edges with very clean ceramic bearings, with no lubrication. I've found this works equally well and without the need to align the two knife edges perfectly horizontal.

 

[skyline1]

One method requires both shaft ends of the rotor to be placed on knife edges which must be perfectly horizontal. The heavy part of the rotor will slowly roll to the bottom.

I use exactly the same method to balance the rotors on my little DeLaval's (see avatar) and have found that, although simple, it works really well

Best Regards Mark

 

[mu38&Bg#]

The phone apps are not doing dynamic (dual plane) balancing.

Static balance is acceptable when the shaft is short or is a single disk.

 

[ddmckee54]

Are you building a refrigeration unit? The system looks similar, though a lot smaller, to the old steam heated chillers that were used in some of the buildings that I worked in - many... many... moons ago. (mid-80's)

 

[ddmckee54]

If you know of any good methods for dynamic balancing, please do let me know.

I remember watching a video of a guy building a small gas turbine. I think he machined his own compressor wheel and then dynamically balanced it. If it's the guy I'm thinking of, he builds some fairly high-tech projects and then gets into the science and engineering behind them. I'll see if I can find it again.

 

[HMEL]

Thanks for the compliments on the machine work,...it's all done on my DIY CNC Lathe and Mill. I think it's been about 8 or 9 years ago that I built my CNC Milling machine, and with that and a CNC rotor tool I also built, I'm able to machine most everything I need, including those internal splines. Here's a link to a YouTube video I made months ago on how I did it: Cutting Internal Splines.

You're right about this project not being an impulse build, the engine has a very specific purpose. Most folks look at the hot flames coming out of a jet engine and believe it's all those hot gases that make the thrust, and that is true, but a jet engine's exhaust doesn't need to be hot in order to produce thrust, the gases just need to be moving and have mass. F=ma governs how much static thrust a jet engine produces; the mass of the exhaust gases times it's acceleration is how thrust is calculated,....and it's the compressor inside the engine that is doing all the work to produce that thrust. Essentially, jet engines are just big air compressors that direct their compressed air out of the back of the engine.

It really doesn't matter what type of motor drives a jet engine's compressor, as long as that motor has adequate power and can spin the compressor at the required rpm the compressor will produce thrust.

So that's my first step; determine how much thrust a steam turbine can produce. If all goes as I hope it will, that's the first step in an even larger project.

What you have in mind to build is called a combined cycle plant. Now as a technical point the jet engine compressor is designed to take the working fluid (air ) to a higher Pressure in a combustion can. Here fuel is added and the combustion products with even higher pressure are exhausted out the exit nozzle which yields a thrust. If the compressor is driven by a motor you do not need the driving blades on the shaft to power the compressor. Almost 70% of a turbine developed power drives the compressor. Its a pressure drop that determines the power. All of this can be calculated. In a combined cycle the hot gas from the turbine is dropped into the furnace of the steam boiler at about 1500 degree F. To calculate the steam turbine take the inlet pressure and calculate the drop in the condenser and that yields the horsepower theoretical number. The actual number will be less by some efficiency number. A basic textbook in thermodynamics will demonstrate the proper technique to determine what is called Pressure Volume work. For examples you might look at the guys who fly model airplanes with jet engines. They will not have the steam cycle attached but they have everything else. Its actually amazing how small they can make these things.

 

[Toymaker]

Are you building a refrigeration unit? The system looks similar, though a lot smaller, to the old steam heated chillers that were used in some of the buildings that I worked in - many... many... moons ago. (mid-80's)

No, the system is not for refrigeration. Once the turbine, boiler, condenser, etc are all working well together I will connect a centrifugal compressor (the same you'll find on automotive turbo chargers) directly to the turbine's output shaft which will compress large volumes of air to about 50 to 60 psi.

 

[Toymaker]

Now as a technical point the jet engine compressor is designed to take the working fluid (air ) to a higher Pressure in a combustion can. Here fuel is added and the combustion products with even higher pressure are exhausted out the exit nozzle which yields a thrust.

Your explanation of how a jet engine works is mostly correct, but with one small mistake: during normal operation, the combustion chamber and the combustion gases are always at a slightly lower pressure then the compressor outlet. The temperature of the combustions gases are clearly much higher, but the pressure is always a few psi lower. When pressures inside the combustion chamber or it's exhaust inadvertently reach a higher pressure then the compressor outlet the combustion gases flow backwards through the compressor and is one cause of what's known as a "compressor stall".

If the compressor is driven by a motor you do not need the driving blades on the shaft to power the compressor. Almost 70% of a turbine developed power drives the compressor. Its a pressure drop that determines the power. All of this can be calculated. In a combined cycle the hot gas from the turbine is dropped into the furnace of the steam boiler at about 1500 degree F. To calculate the steam turbine take the inlet pressure and calculate the drop in the condenser and that yields the horsepower theoretical number. The actual number will be less by some efficiency number. A basic textbook in thermodynamics will demonstrate the proper technique to determine what is called Pressure Volume work. For examples you might look at the guys who fly model airplanes with jet engines. They will not have the steam cycle attached but they have everything else. Its actually amazing how small they can make these things.

I'm not using the exhaust of a small jet engine as my heat source, though I do admit that the burner exhaust in the video I posted above does look and sound a lot like a jet engine, it is just a diesel fuel burner with a "leaf blower" forcing high speed air through the burner.

I have not yet measured the exhaust gas temperature of my burner but since I'm not adding any cooling air to the exhaust, as is done with all jet engines, I suspect Adiabatic flame temperatures of around 2100C (3800F)

 

[Steamchick]

Hi Toymaker. An impressive project. I am trying to understand your boiler concept.... it looks like a flash boiler? Liquid injected at one end and superheated vapour extracted at the other? One point to research.... you may have done it already? I.E. The corrosion resistance of aluminium at the expected max temperature of your organic fluid. Please can you post more drawings of the boiler?
K2

 

[Steamchick]

Hi Toymaker, I was looking at bearings - considering something for a turbine that I have some parts and ideas about. I read about these bearings:
KLNJ R Series 2RS C3 Rubber Sealed High Speed Imperial Bearings - High Quality
"High speed" means 35,000rpm. for class C3. Apparently they are "slack" to permit the balls to expand when they get hot at speed...
So for higher speeds, we need special bearings, that I have not found yet.
What class/material of bearings do you plan to use? ( - for planned 70,000rpm max speed? (Your turbine is bigger than I am looking at). I'm sure it is in the post - all I have seen is "full ceramic thrust bearing..." ).
Ta,
K2.

 

[Toymaker]

Hi Toymaker. An impressive project. I am trying to understand your boiler concept.... it looks like a flash boiler? Liquid injected at one end and superheated vapour extracted at the other?

I have struggled to draw a 3D approximation of what my boiler should look like, and although I've managed to sketch out a few pencil and paper drawings, they're all so bad that I'm the only person that can interpret them. The best I can do for now is to describe the 2D drawing below.

All the small green circles represent 5/16 (8mm) OD boiler tubes, except that I got lazy and didn't show all of them, so in the areas you see green circles, imagine that space is filled with boiler tubes. Flames from the burner are forced to flow around the tubes turning 180 degrees and flow out the Exhaust. Sheet metal baffles not shown will direct the combustion gases to take a spiral path around the outside of the burner, thereby maximizing the time the gases are in contact with the boiler tubes.

The boiler starts out as a mono-tube boiler having a single inlet tube for high pressure Freon from the Boiler Feed Pump, but soon branches out into 2 tubes, then 4 tubes, and finally into the 8 separate tubes which directly connect to the turbine. Many small tubes provide a much larger surface area and smaller diameter, both of which increase heat transfer from the combustion gases to the "steam". This drawing is a "schematic", it is not to scale, nor does it show the actual placement of the boiler tubes.

Do we call this a "Flash Boiler"? Some will say yes, others will say no,...I'll let the reader decide.
It will however be a Supercritical Boiler as I intend to keep pressure and temperature nearest the boiler outlet just above the critical point of the working fluid, thereby keeping the working fluid in a liquid state until it exits the steam nozzles inside the turbine. This serves several purposes: liquids conduct heat better than gases, so by keeping the working fluid in it's liquid state inside the boiler tubes, the fluid is able to absorb heat from the combustion gases faster and transport that heat out of the boiler faster. Second, the boiler tubes are Aluminum and will quickly be melted by the combustion gases if they are not kept cool by the working fluid carrying away the heat quickly enough. Finally, supplying the convergent-divergent nozzles with a liquid on the supply side allows that liquid to flash into a vapor on the exit side resulting in the highest possible gas velocity flowing into the turbine blades.

One point to research.... you may have done it already? I.E. The corrosion resistance of aluminium at the expected max temperature of your organic fluid. Please can you post more drawings of the boiler?
K2

Thanks for the safety advice,...it's always welcome.
As much as possible, I'm trying to chose a working fluid that isn't corrosive to aluminum at higher temperatures. My current first choice is R-123, which has been used successfully in other ORC engines. However, the critical point is 184C and at temperatures above 250C it can decompose into hydrochloric acid, hydrofluoric acid, and carbonyl halides. Since the combustion gases will likely be at 2100C (3800F) I do not know if any of the R-123 molecules will be decomposed as they contact the tube's inner walls. Unless someone here knows the answer, this is something I will need to determine experimentally.

 

[Toymaker]

Hi Toymaker, I was looking at bearings - considering something for a turbine that I have some parts and ideas about. I read about these bearings:
KLNJ R Series 2RS C3 Rubber Sealed High Speed Imperial Bearings - High Quality
"High speed" means 35,000rpm. for class C3. Apparently they are "slack" to permit the balls to expand when they get hot at speed...
So for higher speeds, we need special bearings, that I have not found yet.
What class/material of bearings do you plan to use? ( - for planned 70,000rpm max speed? (Your turbine is bigger than I am looking at). I'm sure it is in the post - all I have seen is "full ceramic thrust bearing..." ).
Ta,
K2.

I'm using a 6000 (10 x 26 x 8mm) on the front end of the turbine shaft and a 6201 (12 x 32 x 10mm) on the back end of the shaft. Both are full ceramic bearings and are made by "Mochu", which I believe is a Chinese company. Neither bearing was overly expensive so I doubt they're the very best quality, but at least until I'm more confident that my ORC engine has good potential, I'm unwilling to invest $150 per bearing for ABEC 7 hi quality bearings.

I'm not familiar with the term, "class C3". What I see used by bearing retailers and manufactures is the ABEC scoring system, where ABEC 1 is the lowest quality and ABEC 7 is the highest.

What I learned about ceramic bearings while I was still in the working world has left me fairly confident with using even lower quality ceramic bearings. Some 15 years ago Lockheed Martin hired an outside company specializing in long term failure analysis; their job was to continuously spin a rather heavy gyroscope while putting it through extremely hot & cold cycles. They did this for 6 months, at which time the bearings were disassembled and inspected for signs of wear. There was absolutely zero wear,...and the bearings had been run the entire 6 months with no lubrication,...they were run dry. So as long as the bearings I have are of reasonably good quality and don't induce unwanted vibration, I'm rather confident they'll last for a very long time, even at RPMs well beyond their rating.

Also, don't flood your hi rpm bearings with lots of oil, instead find a way to spray the bearing with an oil mist.