Ambitious ORC Turbine

[Toymaker]

I am fascinated by your project. Considering the exhaust from your burner: and the coanda effect nozzles....
The momentum exchange of moving gases reacting on the host vehicle causes thrust. So if the host vehicle generates a lot of low pressure hot gas, made of larger molecules than N2 and O2 in air, then it may be a good idea to inject this gas into the coanda nozzle. The basis is that this gas is initially a part of the vehicle, so when accelerated and ejected from the nozzle will add to the momentum ejected backwards, thus increasing the momentum reaction force forwards....
I think?
If so, perhaps it should be in the form of a De Laval nozzle in the middle of the coanda nozzle, and thus the exhaust gas jet can enhance the mass of air drawn through the coanda nozzle, utilising the last of the heat energy in the exhaust as it expands and cools adiabatically through the nozzle...?
Sorry if this is a crazy idea, but perhaps worth considering?
K2

Hmmmm, that's an interesting idea. As the exhaust gases exit the boiler I could easily rout them into the intake of the large centrifugal compressor, where they would be mixed with a much larger volume of ambient air.

Your idea of using a De Laval nozzle to shoot high velocity air into the middle of the Coanda nozzle is very similar to what was done on the XFV-12 program; below is a diagram of the Augmenter Geometry they used. The structure in the top middle of the diagram is a nozzle.

If you look closely at the first diagram I posted, you'll see a similar set of shapes as the one above, in the upper right portion of my drawing.

 

[Toymaker]

I think (again, not an expert, just a novice learning how to do this) - that you can take the 150kW and translate it into the kgs per second of liquid HCFC123 you need to pump into the boiler to extract the heat, by using the difference in Enthalpy of the HP and cold liquid (post pump) to the Enthalpy of the HP and hot liquid. - Values taken form the table I attached earlier - as the pump power (I think?) does the change of state from LP cold gas to HP liquid...?
Ta,
K2

When I did the rough power output calculations for the turbine, I started by determining the maximum mass of R123 that would be flowing through the nozzle. The nozzle is the most limiting part of the overall engine.

Gas flow rates through a nozzle cannot exceed the local speed of sound, which for R-123 at the critical point is 78.1 m/s. Total nozzle area is 0.231 sqr in or 0.000149 sqr meters. So, 78.1 m/s x 0.000149 sqr meters = 0.0116 cubic meters/sec. From our favorite thermo properties table, I find the density of R123 at the critical point to be 449 kg/cubic meter, and since I only have 0.0116 cubic meters, my total mass flow is 5.2 kg/s.

 

[Steamchick]

Re: post #81: I think the feeding of the compressor with exhaust gas will have 1 advantage of increasing the density of the compressor gas. But also possibly higher pressure and flow due to pressurising the inlet if the inlet with injected exhaust gas is design as a "gas injector", to draw-in extra air. I make gas burners for small model boilers, and induced air from the gas jet requires a particular geometry (not unlike a Delaval nozzle). I.E. A converging throat to venturi with controlled expansion following. This could sort of supercharge the compressor.... It is all about pressure change to velocity to develop a sub-atmospheric pressure gradient that sucks-in extra air, then slows the mixture to re-develop pressure above atmospheric....
I knew a guy (50 years ago) who increased his engine power on his drag racer by "exhaust gas injection" to induce extra air into the carburettor. It was a permitted "supercharging", as there was no mechanical device forcing the draught. It only worked as the carb was fully open for the drag.... and was re-jetted accordingly.
K2

 

[Steamchick]

I am enjoying this. The Engineering involved : sizing the fuel pump (post #80), then "energy management" (heat flow) through the rest of the system, is a real education for me.
Thanks Toymaker!
You can write the book when you finish - a text book example of system design!
K2

 

[Toymaker]

I am enjoying this. The Engineering involved : sizing the fuel pump (post #80), then "energy management" (heat flow) through the rest of the system, is a real education for me.
Thanks Toymaker!
You can write the book when you finish - a text book example of system design!
K2

Thanks for the kind words.

There's a pretty good tutorial on how to calculate power output of a two stage steam turbine using the steam tables at: Engineers Academy The speaker does a good job of explaining how enthalpy drop across each stage determines the power output of that stage. My only complaint is that the example used gives the pressures at both stage outputs, which are unknown values during the initial design. Maybe that part is explained in a different video I haven't found yet.

 

[HMEL]

Actually the power output is a criteria of the original design specification. It starts with a mass flow diagram and the pressure and thermodynamic states are laid out to develop a theoretical horsepower or kw/hr design point. With the turbine itself it starts at the nozzle block control valve then progress downstream to the condenser. Use of expansion curves and consideration of nozzle design are set before the wheels and blades are assembled. If extraction points are required as in the case of feed water heaters it is even more important. If under contract and if the stakes are high as in large utility machines these will be measured and tested to see if they meet specs. Same for the boiler The manufacture or designer will usually over design to meet this purchase spec called manufactures margin. So you do not normally build a turbine and see what you get. The design number will determine the heat rate on how much energy per kw is produced. For smaller turbines probably not as important. The manufacture will know or should know what the pressure drop his stages will be. This will depend upon the type of blade impulse , reaction or hybrid of both types. This design process is used on all systems no matter the working fluid Steam, Ammonia, or Organic Fluids

And buried under all this theory is the practical aspect of how these things are made. You do not wish to have the blades leave the wheel due to nozzle passing frequency issues, bad root design or over speed. Over the years I have kept about 8 of these types of blade failures just as souvenirs of interesting things that can happen.

The average modeler probably doesn't care about all the engineering that goes into it. If it spins at any rpm and makes a little power he is probably a happy camper.

 

[Steamchick]

Thanks HMEL. Common sense from someone with experience!
My experience doesn't include turbines, so I am a beginner in the learning process, which is why I have been asking lots of questions and challenging Toymaker (a man of great patience!). But I come from a professional background, where we calculated everything before making it. Much easier to change the pencil-on-drawing before cutting metal, than trying to get the metal to do what it cannot! This methodology was new when I joined a certain design office, as most of the work was "drawn first". Then "made and tested and re-designed to fix the faults"..!!! As long as people followed the 3 LARGE BOOKS of the "Design manual" (a list of designs that had worked!) then they didn't need calculations (Ha! Ha!). I joined, having come from a different company, and started design by planning the job, doing calculations on existing designs, then changing to improve to get the required performance of my new designs. First parts: The steel fabrication shop complained because I used "standard steel structural designs" and "it was too awkward to set-up for welding", then "much to light - it will collapse" criticism. Neither of which was true. And after 4 years I left (the writing was on the wall! - "They have been weighed in the balance and found wanting") as they were closing the place down, except for my designs and a couple of others! So I feel justified in blowing my own trumpet as their walls came tumbling down... "New gates" cannot save the city if the walls are collapsing... Here-endeth the lesson..
K2

 

[Toymaker]

Actually the power output is a criteria of the original design specification. It starts with a mass flow diagram and the pressure and thermodynamic states are laid out to develop a theoretical horsepower or kw/hr design point. With the turbine itself it starts at the nozzle block control valve then progress downstream to the condenser. Use of expansion curves and consideration of nozzle design are set before the wheels and blades are assembled. If extraction points are required as in the case of feed water heaters it is even more important. If under contract and if the stakes are high as in large utility machines these will be measured and tested to see if they meet specs. Same for the boiler The manufacture or designer will usually over design to meet this purchase spec called manufactures margin. So you do not normally build a turbine and see what you get. The design number will determine the heat rate on how much energy per kw is produced. For smaller turbines probably not as important. The manufacture will know or should know what the pressure drop his stages will be. This will depend upon the type of blade impulse , reaction or hybrid of both types. This design process is used on all systems no matter the working fluid Steam, Ammonia, or Organic Fluids

And buried under all this theory is the practical aspect of how these things are made. You do not wish to have the blades leave the wheel due to nozzle passing frequency issues, bad root design or over speed. Over the years I have kept about 8 of these types of blade failures just as souvenirs of interesting things that can happen.

The average modeler probably doesn't care about all the engineering that goes into it. If it spins at any rpm and makes a little power he is probably a happy camper.

Question:
When calculating the power developed by each turbine stage using the change in enthalpy across each stage, as was done in this Engineers Academy example, does the term "stage" include both stator and rotor blades, or just the rotor blades? Since the job of the stator blades is to re-direct the gas flow, and these blades are not moving, the enthalpy drop across the stators results in increasing the steam velocity which will result in greater energy transfer onto the adjacent rotor. So it seems logical that I should include the enthalpy drop across both stator and rotor blades to find the total power for each turbine stage. Or, since the stator is not moving, it therefore does not directly contribute any power transferred onto the rotor blades and the enthalpy drop across the stator should be excluded.

Which is the correct answer ?

 

[Steamchick]

I don't know, but would guess the enthalpy drops across the static and dynamic pair....?
K2

 

[Toymaker]

I started working on the ECU (Engine Control Unit) (aka: FADEC) a few days ago; I have been ordering various parts for months, and I finally sat down and wired the major electronic control boards to the micro controller (Arduino MEGA 2560) and a few other parts, such as the air blower (leaf blower) for the burner, fuel pump (wobble plate air compressor) and High voltage ignitor board. I have a nice Aluminum Mini ITX Computer Case on order through eBay to put all the boards in, but for now, they're all just laying out on my desk; it's a nice little rat's nest .

The little 4" color display shown in the video displays various parameters such as boiler input & output temperatures, and pressures, fuel ignitor status, burner flame sensor, fuel pump & air blower power levels, and of course, turbine rpm.

All the variable resistors mounted in the clear-plastic box make up the test rig, which allows me simulate various conditions such as over-temperature or pressure, or a feed-pump failure, without the risk of forcing these conditions in the real engine. The test rig allows me to check the software I'm writing to make sure it keeps the engine running within safe operational limits, before I give the computer full control.

Once I'm able to mount all the boards into that nice little aluminum box I ordered, the one that still hasn't arrived yet, I will be able to carry all the electronics outdoors where I will begin testing various software routines such as "start-up", and "power level".

 

[Bentwings]

very interesting Rc model jets have something similar most use an electric motor to get the turbine spinning then automatically turn on the fuel pump an monitors temp once it’s up to operating temp and speed a light comes on that signals the modeler that it’s ready for takeoff power . Most have in flight automatic monitoring to prevent over speed with either auto slow down or speed limiting Basically you plug in the starting unit or motor and push the button . These have evolved over the years after many fires and runaways I actually only know second hand as jets are far above my pay scale LOL
I started working on the ECU (Engine Control Unit) (aka: FADEC) a few days ago; I have been ordering various parts for months, and I finally sat down and wired the major electronic control boards to the micro controller (Arduino MEGA 2560) and a few other parts, such as the air blower (leaf blower) for the burner, fuel pump (wobble plate air compressor) and High voltage ignitor board. I have a nice Aluminum Mini ITX Computer Case on order through eBay to put all the boards in, but for now, they're all just laying out on my desk; it's a nice little rat's nest .

View attachment 145998

The little 4" color display shown in the video displays various parameters such as boiler input & output temperatures, and pressures, fuel ignitor status, burner flame sensor, fuel pump & air blower power levels, and of course, turbine rpm.

All the variable resistors mounted in the clear-plastic box make up the test rig, which allows me simulate various conditions such as over-temperature or pressure, or a feed-pump failure, without the risk of forcing these conditions in the real engine. The test rig allows me to check the software I'm writing to make sure it keeps the engine running within safe operational limits, before I give the computer full control.

Once I'm able to mount all the boards into that nice little aluminum box I ordered, the one that still hasn't arrived yet, I will be able to carry all the electronics outdoors where I will begin testing various software routines such as "start-up", and "power level".
View attachment 145999

 

[olf20]

I follow with great interest. I marvel at yours and
other peoples abilities and just plain brain power.
As a dabbler in Arduino's what display are you using
on your ECU?
Keep up the project I hope to see your first demo!
olf20 / Bob

 

[Toymaker]

I follow with great interest. I marvel at yours and
other peoples abilities and just plain brain power.
As a dabbler in Arduino's what display are you using
on your ECU?
Keep up the project I hope to see your first demo!
olf20 / Bob

Hello Bob, the display is a 4.0" Inch 480x320 TFT LCD, ST7796S SPI with Touch. I purchased the display on eBay from this seller: modul_technik Data sheets and demo software can be found here: LCDWIKI
This display uses the 4-wire SPI interface to communicate, and has Touch-Screen capability, but I'm not using that feature.

 

[HMEL]

Question:
When calculating the power developed by each turbine stage using the change in enthalpy across each stage, as was done in this Engineers Academy example, does the term "stage" include both stator and rotor blades, or just the rotor blades? Since the job of the stator blades is to re-direct the gas flow, and these blades are not moving, the enthalpy drop across the stators results in increasing the steam velocity which will result in greater energy transfer onto the adjacent rotor. So it seems logical that I should include the enthalpy drop across both stator and rotor blades to find the total power for each turbine stage. Or, since the stator is not moving, it therefore does not directly contribute any power transferred onto the rotor blades and the enthalpy drop across the stator should be excluded.

Which is the correct answer ?

In general the stage is ahead of the nozzle blocks. These are designed to increase the velocity of the flow but there will be a penalty. The rotor blades are designed to accept the high velocity and extract momentum energy. It is pressure volume work. But you should also know there are two types of nozzle blocks, impulse and momentum. The math is similar. The power delivered depends on the turbine back pressure. With good design that is.

 

[Steamchick]

Hi Bob (olf20) - I suggest you do not include me in the "Clever Botherers" list... I am not as clever as the words I use... I.E. I waffle a lot, but I am not sure if what I am proposing actually makes sense, just trying to prompt others to generate definitive answers that I can learn from! HMEL, Toymaker, 'n all are the Clever Botherers!
K2

 

[Toymaker]

Managed a bit more work on the ECU. Turned out I needed 2 Mini iTX Cases to fit all the electronics. Anyway, here's a first look under the hood,...still have lots of wires to solder in

 

[Toymaker]

ECU Status Update:
Much has happened since my last report from 3 months ago; the biggest change was my finally accepting the need for a larger electronics box to put all the various boards into. The two mini iTX boxes shown my previous post have been replaced by a single, decades old, mini-tower chassis, shown below.


I also moved the two High Voltage boards into their own small metal box to give better shielding from the electrical noise they generate. The output from both boards are in series which doubles the voltage allowing for a longer spark gap. The two boards run cool which allows me to keep them operating continuously during testing, however, during normal operation, a flame sensor will allow the computer to turn off the ignition arc as soon as the fuel has been ignited.




The above spark is approximately 1 cm long and ignites the diesel mist very effectively.

Finally, I have the software working well enough that I can now take all the hardware outside to my test stand and determine best start-up parameters, ie how much fuel and how much air flow.

 

[Toymaker]

Finally, the Burner is now under computer control ECU & Burner test Video

The pic below shows my outdoor test stand. The one liter bottle of fuel shown lower right is B7 diesel.

In the video, after I press the start button, about 3 seconds was required to suck fuel into the burner and ignite. The flame is detected and the ignitor is turned off. At 19 seconds into the video an unexpected air bubble in the fuel line caused the flame to go out, which the sensor instantly noticed resulting in the ignitor being turned on until the fuel is re-ignited, followed by flame detection and ignitor once again being turned off. That sequence took only 1 second, for which I'm very pleased as I was concerned that a flame-out might possibly result in a cloud of unburned fuel mist blown into the boiler where it could explode or at least result in an unpleasant pop.
Almost immediately after the re-ignite cycle, the ECU increases fuel and air flow, as it was programed to do, in order to increase boiler heating. The increased power output continues until I turn the “Pressure Out” up to a setting above 20 (an arbitrary number chosen for test purposes only). At this point the ECU regulates the power output based on the manual input, “Power Request”, which I turn up to about 23% before I press the off button.

After the start-stop button is pressed again, fuel flow is turned off while the air blower remains on at 10% to blow excess heat out of the combustion chamber, which is done to prevent the plastic flame sensor from melting due to latent heat within the metal parts of the combustion chamber.

Next steps, before I place the burner inside the boiler: wire up the boiler feed pump (a pressure washer) and write the software to control it.

 

[Toymaker]

Adjusted burner air & fuel flow to maximize Blue Flame at higher fuel burn rates. I haven't been able to get a blue flame at idle burn rate; and I'm not overly concerned about idle burns as I don't expect the boiler will spend much time at idle. The burner is reduced to idle just before it's turned off and clearly shows the yellow-orange flames. This video was taken at night with limited outdoor lighting. I posted a longer video on YouTube Night Burner Test

 

[Steamchick]

Just watched the longer version. It may have been more interesting if you had added a commentary of "50% power", "full power" etc. per burn test. I guess that is what it was showing? : Set blower speed, adjust fuel to optimum flame, record settings....
What a great burner! You have spent a LOT of time developing this and have now got something you can reliably turn ON and OFF and adjust the power. Ideal for your boiler!
Well done!
K2