Adding an ejector component?

[timorichert]

Dear all,

I have started having a look at TESPy (and I'm impressed!), and also the heatpumps repository, offering readily modelled heat pump topologies based on TESPy (https://github.com/jfreissmann/heatpumps). Many heat pumps process improvements (IHX, economizer, cascading, ...) are covered by this excellent combined work. I'm wondering if it would be possible to also model an ejector as a heat pump process improvement, allowing to recover some of the otherwise lost work thrown away in the throttling valve.

Ejector Cycle example (Source: Annex 58 High-Temperature Heat Pumps Task 1 – Technologies, Task Report, https://heatpumpingtechnologies.org/publications/annex-58-high-temperature-heat-pumps-task-1-technologies-task-report/)

Here is an example of a Python code repo that models an ejector, and that could be used outright or offer guidance how it could be implemented: https://github.com/AdamBuruzs/simpy_ejector

Happy to hear your thoughts!

All the best,
Timo

P.S.: I had searched for ejector before writing, and hope I didn't miss a previous discussion!

[fwitte]

Hi @timorichert and thank you for reaching out!

There was no public discussion I am aware of, thank you for bringing it up. I actually discussed this topic already with two groups of people, one of them being the main authors to the heatpumps dashboard.

I will have to take a look on how the python package you linked here is modeling the component, and how to transfer that to tespy. There is quite a lot of things going on right now, so I cannot promise a quick implementation. If you would be willing to help out here please let me know. Maybe we can form a small task-force with the other interested people, looking at you @jfreissmann and @maltefritz. I could offer a little bit of guidance on how to implement new components to accelerate the first version. I will update you, once I find the time to look into that library.

Best

Francesco

[timorichert]

Hi Francesco,

thanks for your prompt reply! Good to hear that this topic is already under discussion.

Your link looks promising as well. I guess calculation speed will also be relevant, since TESPy is rather quick, and an overly complex component might slow the process down. As for simpy_ejector, I have posted a comment on a related issue on implementing an interface with CoolProp as well: AdamBuruzs/simpy_ejector#1.

Unfortunately, I cannot dive into this at the moment, but I'll see if I can tackle that in the near future.

All the best,
Timo

[jfreissmann]

Hi @timorichert, hi @fwitte,

thank you for bringing up the ejector component. @maltefritz and I view it as the biggest missing piece of heatpumps as well and would love to help getting it to tespy and eventually heatpumps. In addition to the resources you two brought up, the C# library portyanikhin/VCRC also models ejectors (see e.g. EjectorFlows.cs). This could be another point to work from, but unfortunately there are no refereces to literature given that I was able to find.

[jfreissmann]

Furthermore, there is a paper with the title Thermodynamic modelling and parameter determination of ejector for ejection refrigeration systems that looks extensive but promising.

[fwitte]

The description of the model actually looks quite straight forward... Sound of speed is also available in CoolProp. One could give it a shot! Open questions could be the estimation of correct dimensions of the ejector parts, and the internal iteration in the simulation process description might create a little bit of overhead. It should be doable nevertheless :)

[fwitte]

I started a notebook over here: https://github.com/fwitte/ejector (only first two steps, and the second step has a bug, which I could not find very fast). You are all invited to have a look :). Later the model should be solved in one step with a large system of equations. Potentially, the ejector could be a subsystem in tespy using existing components.