Gas cycle, rankine cycle, organic rankine cycle, kalina cycle, and all that

[lalbatros]

(maybe it fits better in this forum)

Hello,

I have been asked an advice about waste heat to power systems (say wh2p).
The available waste heat to be considered is a (dirty) gas in a temperature range between 250°C and 450°C.
The useful power that could be generated would range between 2MW and 10MW.

I know some examples based on steam turbines, either the rankine cycle or the organic rankine cycle.
I have read quite a lot about that, but I not a specialist of this topic.
I am interrested by answers to some very naïve questions like these:

• Apparently the back pressure in wh2p steam turbines is above athmospheric pressure. Is that right? And what about the temperature at the turbine outlet?
• If this is right, why is that so? Going to lower pressure could increase the efficiency, isn't it? Is the cost too high? Why? Is there any other drawback?
• I have never seen any application based on a gas cycle? Why?
• What is the advantage of a steam-based cycle conmpared to a gas-cycle in such circumstances?
• A gas cycle could be build with a compressor, and heat exchanger and a gas turbine. I think this is called "indirect cycle" as compared to a gas cycle based on a combustion chamber. Have such devices been used? If no, why?
• Naïvely, I think that the lowest temperature in a cycle is the most important parameter. Is that true?
• From all the cycles mentioned in the title, I think that none have any special advantage or disadvantage, as far as the lowest temperature in the cycle is concerned. What do you think?
• The overall efficiency of a wh2p plant results from the heat-to-power efficiency as well as from the heat exchanger efficiency. This exchanger is needed to bring heat from the waste stream to power stream. Therefore, focusing of the cycle efficiency is not enough and could even lead to wrong conclusions.
• How can the "organic rankine cycle" (orc) and the "kalina cycle" (kc) be compared?
  (orc is based on a pure fluid with low boiling point, kc is based on an ammoniac-water mixture with a variable boiling point)
• In other words, when comparing orc and kc performance, how to avoid comparing apples and pears? Is it only possible to draw general comparisons?
• As far as the cycles efficiency are involved, I would guess that orc and kc have the same efficiency if the boiling point is the same. Do you agree?
• Is the Carnot efficiency based on max and min temperatures practically relevant to compare the orc and kc cycles? Or would there be obvious big differences between orc and kc that would make their cycle efficiency obviously and significantly different?
• The different fluids (pure organic versus ammoniac/water mixture) could make a difference on the heat exchanger side. Is it possible to draw a general conclusions, or is it necessary to study specific cases?
• For sure, the kc has one additional degree of freedom: the mixture level. Therefore, it should be no surprise that an optimisation would be possible, leading to an advantage over the ocr cycle. What do you think about that?
• Does that mean that designing a kalina cycle implies optimizing the cycle (the mixture) for the given waste heat stream in order to get a real advantage over another cycle?
• Would that also mean that any deviation from the optimum (either in implementation or in operation) could just vanish the benefit of the kc?
• Would there be a tendency to complicate and over-design a kalina cycle and impact the cost negatively?
• If the simplest version of the kc is chosen, what are the degree of freedom to optimize the overall efficieny? I see one parameter: the ammoniac/water mixture on the colde side that determines the lowest cycle temperature. Would there be a second or a third parameter?
• If there is only one parameter to optimize in the simplest kalina cycle, why would the kalina efficiency be (always) better that the orc efficiency for the same temperatures?
• The advantages of the orc and kc decrease as the hot side temperature increases? Do you agree? Would you have some data about that?
• Would have some data or ideas about the costs involved?
• When does it make sense to consider orc?
• When does it make sense to consider kc?
  
• Last but not least, any comment about wh2p is of interrest. Specially comment that could help a non-specialist to sort out what is important and what is not.

With my thanks and best regards,

Michel

 

[squil]

Regarding the ORC cycle, its use, and its optimization in WHR applications, you can read that paper:
http://orbi.ulg.ac.be/handle/2268/14609

I don't have any reference about the Kalina cycle, but I know that, although its overall efficiency can by slightly higher, its cost is also much higher due its complexity.