Under-determined System with Compressor Maps & MovingBoundaryHeatExchanger Offdesign

[lennipapa]

Hi,

I'm implementing part-load analysis for the two-stage NH3 heat pump from the documentation and have two questions:

1. When using [offdesign=['char_map_pr_group', 'char_map_eta_s_group'], my system is under-determined. What additional constraints are needed for proper compressor map usage in offdesign?

2. For MovingBoundaryHeatExchanger, is there a simpler offdesign method than UA_cecchinato? The required parameters (alpha_ratio, re_exp_r, re_exp_sf) are hard to find in literature and probably differ for different fluids/geometries. Is there a chance to use UA_char similar to regular HeatExchanger?

Thanks!
stepwise (3).py

[fwitte]

Hi,

Hi @lennipapa,

1. When using [offdesign=['char_map_pr_group', 'char_map_eta_s_group'], my system is under-determined. What additional constraints are needed for proper compressor map usage in offdesign?

I guess, you need igva specified as well. You can see from the table of the components, what things you need to specify simultaneously in case of grouped parameter equations: https://tespy.readthedocs.io/en/main/building_blocks/components.html#modules-components-label. Since you are implementing a heat pump, do you mean to implement TurboCompressor characteristics? There is also a PolynomialCompressor class, which you can feed EN 12900 polynomials or measurement data to. There is an example in the docs: https://tespy.readthedocs.io/en/main/api/tespy.components.displacementmachinery.html#tespy.components.displacementmachinery.polynomial_compressor.PolynomialCompressor.

2. For MovingBoundaryHeatExchanger, is there a simpler offdesign method than UA_cecchinato? The required parameters (alpha_ratio, re_exp_r, re_exp_sf) are hard to find in literature and probably differ for different fluids/geometries. Is there a chance to use UA_char similar to regular HeatExchanger?

Not yet, I might implement something soon but cannot promise, or you can make a pull request.

[fwitte]

Hi @lennipapa,

I did not have time to thoroughly look into the issue, were you able to resolve it? In any case, the problem is, that you cannot have variable inlet guide vane angle for both compressors or you must control one of the pressure ratios of the compressors:

cp1.set_attr(design=["eta_s"], offdesign=['char_map_pr', 'char_map_eta_s'])  # pressure ratio of cp1 was fixed previously and remains fixed
cp2.set_attr(design=["eta_s"], offdesign=['char_map_pr', 'char_map_eta_s'])
cp1.set_attr(igva='var')
cp2.set_attr(igva='var')

or

cp1.set_attr(design=["eta_s", "pr"], offdesign=['char_map_pr', 'char_map_eta_s'])  # pressure ratio of cp1 was fixed previously and remains fixed
cp2.set_attr(design=["eta_s"], offdesign=['char_map_pr', 'char_map_eta_s'])
cp1.set_attr(igva=0)
cp2.set_attr(igva='var')

For the standard UA offdesign characteristics you will have to wait a good more month for me to implement it, otherwise look into the heat exchanger source code and modify it to your needs:

tespy/src/tespy/components/heat_exchangers/sectioned.py

Lines 707 to 774 in f6fa9b3

	def UA_cecchinato_func(self):
	r"""
	Method to calculate heat transfer via UA design with modification
	for part load according to :cite:`cecchinato2010`. UA is determined
	over the UA values of the sections of the heat exchanger.
	.. note::
	You need to specify a couple of parameters to use this method. The
	values depend on the context, as they define relations between the
	refrigerant and the secondary fluid. For an evaporator the
	refrigerant is the cold side, for a condenser it is the hot side.
	You can check the linked publication for reference values.
	- alpha_ratio: Ratio of secondary fluid to refrigerant heat transfer
	coefficient
	- area_ratio: Ratio of secondary fluid to refrigerant area
	- re_exp_sf: Reynolds exponent for the secondary fluid mass flow
	- re_exp_r: Reynolds exponent for the refrigerant mass flow
	The modification factor for UA is calculated as follows
	.. math::
	f_\text{UA}=\frac{
	1 + \frac{\alpha_\text{sf}}{\alpha_\text{r}}
	\cdot\frac{A_\text{sf}}{A_\text{r}}
	}{
	\frac{\dot m_\text{sf}}{\dot m_\text{sf,ref}}^{-Re_\text{sf}} +
	\frac{\alpha_\text{sf}}{\alpha_\text{r}}
	\cdot\frac{A_\text{sf}}{A_\text{r}}
	\cdot\frac{\dot m_\text{r}}{\dot m_\text{r,ref}}^{-Re_\text{r}}
	}
	Returns
	-------
	float
	residual value of equation
	.. math::
	0 = UA_\text{ref} \cdot f_\text{UA} - \sum UA_\text{i}
	"""
	alpha_ratio = self.alpha_ratio.val_SI
	area_ratio = self.area_ratio.val_SI
	re_exp_sf = self.re_exp_sf.val_SI
	re_exp_r = self.re_exp_r.val_SI
	refrigerant_index = self.refrigerant_index.val
	if refrigerant_index == 0:
	secondary_index = 1
	else:
	secondary_index = 0
	m_r = self.inl[refrigerant_index].m
	m_ratio_r = m_r.val_SI / m_r.design
	m_sf = self.inl[secondary_index].m
	m_ratio_sf = m_sf.val_SI / m_sf.design
	fUA = (
	(1 + alpha_ratio * area_ratio)
	/ (
	m_ratio_sf ** -re_exp_sf
	+ alpha_ratio * area_ratio * m_ratio_r ** -re_exp_r
	)
	)
	sections = self.calc_sections(False)
	return self.UA.design * fUA - self.calc_UA(sections)

. The equation you want to implement is something like

$UA-UA_\text{design}\cdot\frac{2}{\left(\frac{\dot m_\text{hot,design}}{\dot m_\text{hot}}\right)^\text{re-exp-hot} + \left(\frac{\dot m_\text{cold,design}}{\dot m_\text{cold}}\right)^\text{re-exp-cold}}$

Best

Francesco

[fwitte]

Have you modified the source file in the correct location and does your environment know about that file? How did you install tespy, where is it located and where did you do the modification?

[fwitte]

What does pip show tespy give you as installation location?

[lennipapa]

I modified the TESPy source code on GitHub (https://github.dev/oemof/tespy/blob/f6fa9b395b9475b960bc4ec7d8ebf34cafdf57b1/src/tespy/components/heat_exchangers/sectioned.py#L707-L774), but my local environment is running the official pip version (0.9.13.post2), which probably doesn't have my modifications.

That's my installation location: Location: C:\Users\AppData\Roaming\Python\Python312\site-packages

[fwitte]

Then you need to fork the oemof/tespy repository, clone your fork (git clone https://github.com/lennipapa/tespy) after you created it, uninstall tespy from your environment (pip uninstall tespy) and instead install the cloned version (pip install -e ./path/to/clone/folder). In that clone you can modify the source code and make use of it in your model(s). Also see: https://tespy.readthedocs.io/en/main/community/contribute.html#development-how-label. In the end you can also open a pull request here to insert your changes in the framework :).

[lennipapa]

Thanks, it worked like this! I will do the pull request soon. :)