How to use other fuels in combustion component?

[HaSchneider]

I want to add more fuels to combustion component. Is there a way to do that dynamically with the existing code?
And what is needed for additional fuels? Some needed property might already be provided by coolprop. Is it only the enthalpy of formation defined here?

[fwitte]

Hi @HaSchneider!

That would actually be a great feature! You have to add the formation enthalpy to the location indicated by you, and you have to add the fuel to the fuel list here: https://github.com/oemof/tespy/blob/dev/src/tespy/tools/global_vars.py#L107.

There are two more things to consider:

• The fuels need a lower heating value, which right now is assigned based on the enthalpy of formation, which you identified the correct location. The lower heating value is the calculated based on the stoichiometric combustion, but it only considers C, O and H as reactants.
• Then based on the molar composition the stoichiometry is applied to all fluid components involved with the combustion in the solving process here:

  tespy/src/tespy/components/combustion/base.py

  Lines 527 to 620 in ec1df92

  	inl, outl = self._get_combustion_connections()
  	if fluid in list(self.fuel_list) + [self.co2, self.o2, self.h2o]:
  	###################################################################
  	# molar mass flow for fuel and oxygen
  	n_fuel = {}
  	n_oxy_stoich = {}
  	n_h = 0
  	n_c = 0
  	n_o = 0
  	for f in self.fuel_list:
  	n_fuel[f] = 0
  	for i in inl:
  	n = (
  	i.m.val_SI * i.fluid.val.get(f, 0)
  	/ inl[0].fluid.wrapper[f]._molar_mass
  	)
  	n_fuel[f] += n
  	n_h += n * self.fuels[f]['H']
  	n_c += n * self.fuels[f]['C']
  	n_o += n * self.fuels[f]['O']
  	# stoichiometric oxygen requirement for each fuel
  	n_oxy_stoich[f] = n_fuel[f] * (
  	self.fuels[f]['H'] / 4 + self.fuels[f]['C']
  	- self.fuels[f]['O'] / 2
  	)
  	###################################################################
  	# calculate stoichiometric oxygen
  	n_oxygen_stoich = n_h / 4 + n_c - n_o / 2
  	n_oxygen = 0
  	for i in inl:
  	n_oxygen += (
  	i.m.val_SI
  	* i.fluid.val.get(self.o2, 0)
  	/ inl[0].fluid.wrapper[self.o2]._molar_mass
  	)
  	###################################################################
  	# calculate lambda if not set
  	if not self.lamb.is_set:
  	self.lamb.val_SI = n_oxygen / n_oxygen_stoich
  	###################################################################
  	# calculate excess fuel if lambda is lower than 1
  	if self.lamb.val_SI < 1:
  	n_h_exc = (n_oxygen_stoich - n_oxygen) * 4
  	n_c_exc = (n_oxygen_stoich - n_oxygen)
  	else:
  	n_h_exc = 0
  	n_c_exc = 0
  	###################################################################
  	# equation for carbondioxide
  	if fluid == self.co2:
  	dm = (n_c - n_c_exc) * inl[0].fluid.wrapper[self.co2]._molar_mass
  	###################################################################
  	# equation for water
  	elif fluid == self.h2o:
  	dm = (n_h - n_h_exc) / 2 * inl[0].fluid.wrapper[self.h2o]._molar_mass
  	###################################################################
  	# equation for oxygen
  	elif fluid == self.o2:
  	if self.lamb.val_SI < 1:
  	dm = -n_oxygen * inl[0].fluid.wrapper[self.o2]._molar_mass
  	else:
  	dm = -n_oxygen / self.lamb.val_SI * inl[0].fluid.wrapper[self.o2]._molar_mass
  	###################################################################
  	# equation for fuel
  	elif fluid in self.fuel_list:
  	if self.lamb.val_SI < 1:
  	n_fuel_exc = (
  	-(n_oxygen / n_oxygen_stoich - 1) * n_oxy_stoich[fluid]
  	/ (self.fuels[fluid]['H'] / 4 + self.fuels[fluid]['C'] - self.fuels[fluid]['O'] / 2)
  	)
  	else:
  	n_fuel_exc = 0
  	dm = -(n_fuel[fluid] - n_fuel_exc) * inl[0].fluid.wrapper[fluid]._molar_mass
  	###################################################################
  	# equation for other fluids
  	else:
  	dm = 0
  	res = dm
  	for i in inl:
  	res += i.fluid.val.get(fluid, 0) * i.m.val_SI
  	res -= outl[0].fluid.val.get(fluid, 0) * outl[0].m.val_SI
  	return res

  . That calculates the difference between inflow moles of fluid and the outflow of the same fluid or of its products, which will be zero once converged. I am unsure if this works for fuels that have other components than C, O and H atm. But it could for sure be extended.
• And a small sidenote: this is pressure and temperature independent at the moment, e.g. NOx calculation is would not work right now, but it could be possible to also make the stoichiometry depend on pressure/temperature.

You could give it a try and let me know, how it went?

Best

Francesco

[fwitte]

To make a dynamic method, we could implement an API maybe similar to the pint.Units thing? That class could be having the information on available combustion fuels, their respective enthalpies of formation (or even directly LHV). Then you could add a fuel with some method. The CombustionChamber would pull information from that class once it is prepared for simulation? In contrast to the Units it would also not be a big problem, to have a global COMBUSTION_FUELS instance of that class, which can be manipulated by the user, because it would only be utilized within tespy anyways.

[HaSchneider]

Thanks for the fast reply.
I tried it with acetone and it worked well (as expected as it has only C, O and H). Then I tried it with dicloroethane and it raises the following error.

File \\tespy\src\tespy\components\combustion\base.py:747, in CombustionChamber.lambda_func(self)
    730 def lambda_func(self):
    731     r"""
    732     Calculate the residual for specified lambda.
    733 
   (...)    745             {M_{fluid}}\\ \forall i \in inlets
    746     """
--> 747    return self.calc_lambda() - self.lamb.val_SI

File \\tespy\src\tespy\components\combustion\base.py:798, in CombustionChamber.calc_lambda(self)
    793     n_oxygen += (
    794         i.m.val_SI * i.fluid.val.get(self.o2, 0)
    795         / inl[0].fluid.wrapper[self.o2]._molar_mass
    796     )
    797 n_oxygen_stoich = n_h / 4 + n_c - n_o / 2
--> 798 return n_oxygen / n_oxygen_stoich

ZeroDivisionError: float division by zero

I pushed the changes in my tespy fork.

The suggestion for dynamic method sounds good, I will give it a try later.