SES 36 - WET VAPOUR

[JDPS]

As you can check from previous posts I want to implement coolprop on EBSILON software, in order to use SES36 database. Since this implementation will take me some time, I tried to validate coolprop propertie values for SES36. In order to achive that I used EES with coolprop external routine and solkane 8: http://www.solvaychemicals.com/EN/products/Fluor/Software.aspx. I did a vast number of tests, and conclude that for superheated zone the values obtained from both software are in agreement. Yet for the wet vapor I found huge discrepancies that will influence a cycle simulation, such an ORC. I know that the database for this fluid is based on Monika Thol unpublished paper, yet I would like to inform you that the values doesn´t seem correct. At least considering that the fluid manufacturer (solvay chemical) are correct.

(https://cloud.githubusercontent.com/assets/6802735/2539340/0f76b0ee-b5c7-11e3-9647-962fa206afe9.jpg)

[ibell]

Um, I think you might be reading their density values incorrectly. Things seem fine to me. They (stupidly) give the saturated vapor and liquid in different units. The liquid is in kg/dm^3 and the vapor in kg/m^3. A terrible idea on their part.

Otherwise, what in your opinion is incorrect?

On Thu, Mar 27, 2014 at 4:58 PM, JDPS notifications@github.com wrote:

As you can check from previous posts I want to implement coolprop on EBSILON software, in order to use SES36 database. Since this implementation will take me some time, I tried to validate coolprop propertie values for SES36. In order to achive that I used EES with coolprop external routine and solkane 8: http://www.solvaychemicals.com/EN/products/Fluor/Software.aspx. I did a vast number of tests, and conclude that for superheated zone the values obtained from both software are in agreement. Yet for the wet vapor I found huge discrepancies that will influence a cycle simulation, such an ORC. I know that the database for this fluid is based on Monika Thol unpublished paper, yet I would like to inform you that the values doesn´t seem correct. At least considering that the fluid manufacturer (solvay chemical) are correct.

( https://cloud.githubusercontent.com/assets/6802735/2539340/0f76b0ee-b5c7-11e3-9647-962fa206afe9.jpg )

Reply to this email directly or view it on GitHubhttps://github.com/ibell/coolprop/issues/201 .

[ibell]

Also, there is no guarantee that Solkane 8 properties are correct. If I remember correctly, they just use a cubic EOS, and I think their specific heat values were quite a bit off, which is part of the reason for the problems with the latent heat.

On Thu, Mar 27, 2014 at 5:18 PM, Ian Bell ian.h.bell@gmail.com wrote:

Um, I think you might be reading their density values incorrectly. Things seem fine to me. They (stupidly) give the saturated vapor and liquid in different units. The liquid is in kg/dm^3 and the vapor in kg/m^3. A terrible idea on their part.

Otherwise, what in your opinion is incorrect?

On Thu, Mar 27, 2014 at 4:58 PM, JDPS notifications@github.com wrote:

As you can check from previous posts I want to implement coolprop on EBSILON software, in order to use SES36 database. Since this implementation will take me some time, I tried to validate coolprop propertie values for SES36. In order to achive that I used EES with coolprop external routine and solkane 8: http://www.solvaychemicals.com/EN/products/Fluor/Software.aspx. I did a vast number of tests, and conclude that for superheated zone the values obtained from both software are in agreement. Yet for the wet vapor I found huge discrepancies that will influence a cycle simulation, such an ORC. I know that the database for this fluid is based on Monika Thol unpublished paper, yet I would like to inform you that the values doesn´t seem correct. At least considering that the fluid manufacturer (solvay chemical) are correct.

( https://cloud.githubusercontent.com/assets/6802735/2539340/0f76b0ee-b5c7-11e3-9647-962fa206afe9.jpg )

Reply to this email directly or view it on GitHubhttps://github.com/ibell/coolprop/issues/201 .

[JDPS]

Yes, you´re right regarding the specific mass and volume. Yet, what I think that is wrong is enthalpy variation on the wet vapour phase. This discrepancies will affect the heat exchange on a condenser or evaporator.

[ibell]

The only thing that is relevant is the DIFFERENCE in enthalpy between liquid and vapor. The enthalpies and entropies have different reference states I guess between Solkane and CoolProp. And I think the difference in delta H is due to the specific heat curves.

Ian

On Thu, Mar 27, 2014 at 6:06 PM, JDPS notifications@github.com wrote:

Yes, you´re right regarding the specific mass and volume. Yet, what I think that is wrong is enthalpy variation on the wet vapour phase. This discrepancies will affect the heat exchange on a condenser or evaporator.

Reply to this email directly or view it on GitHubhttps://github.com/ibell/coolprop/issues/201#issuecomment-38831795 .

[JDPS]

I understand the differences on the enthalpy absolute values. Yet, as you stated the enthalpy change is what counts for energy balance, and they are significatively different on both software. I already tested for other fluids using solkane, coolprop and refprop and the differences are neglectable. Even tested on simple ORC simulation, and as expected in the turbine the enthalpy variation is almost the same, yet at the condenser and evaporator they differ a significatively. This discrepancies lead to a cascade effect, since they influence the mass flow values and consequently the power from the turbine. We will have two ORC running on SES36, and I hope that I will be able to get some experimental values in order to compare with coolprop and solkane. Thank you for the support.

[ibell]

Seems the error is with Solkane. There is some data in the paper of Riva (http://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=1852&context=iracc), and it seems that CoolProp (M.Thol's EOS) agrees fairly well with the data, but not well with Solkane. You should send them an email, perhaps pointing to this issue. Here is a script that generates the figure below:

import numpy as np, matplotlib.pyplot as plt, CoolProp.CoolProp as CP

#from Riva paper
_T = [283.15, 293.15, 303.15, 313.15, 323.15, 333.15, 343.15, 353.15, 363.15, 373.15, 383.15, 393.15, 403.15, 413.15, 423.15, 433.15, 443.15]
_p = [0.395, 0.579, 0.833, 1.174, 1.622, 2.2, 2.932, 3.845, 4.964, 6.316, 7.929, 9.831, 12.055, 14.636, 17.622, 21.072, 25.067]
_rhoL = [1400.14, 1377.21, 1353.24, 1328.22, 1302.16, 1275.04, 1246.85, 1217.59, 1187.26, 1155.82, 1123.24, 1089.48, 1054.38, 1017.66, 978.55, 934.72, 875.19]
_rhoV = [2.89, 4.12, 5.77, 8, 10.97, 14.9, 20.08, 26.82, 35.45, 46.31, 59.79, 76.45, 97.2, 123.68, 158.94, 209.17, 289.01]
_hL = [207.26, 215.92, 226.05, 237.62, 250.5, 264.36, 278.74, 293.04, 306.69, 319.35, 331.04, 342.12, 353.16, 364.78, 377.61, 392.33, 410.03]
_hV = [359.64, 369.63, 379.7, 389.66, 399.25, 408.23, 416.31, 423.29, 429.05, 433.67, 437.38, 440.51, 443.37, 446.2, 449.1, 451.93, 454.18]
_sL = [1.026, 1.056, 1.09, 1.127, 1.168, 1.21, 1.252, 1.293, 1.331, 1.365, 1.396, 1.424, 1.451, 1.479, 1.509, 1.542, 1.582]
_sV = [1.564, 1.58, 1.597, 1.613, 1.628, 1.642, 1.653, 1.662, 1.668, 1.671, 1.673, 1.674, 1.675, 1.676, 1.678, 1.68, 1.681]

for T in [35.6, 142, 47]:
    print T, (CP.PropsSI('H','T',T+273.15,'Q',1,'SES36')-CP.PropsSI('H','T',T+273.15,'Q',0,'SES36'))/1000

_hfg = (np.array(_hV)-np.array(_hL))*1000

hfg = CP.PropsSI('H','T',_T,'Q',1,'SES36')-CP.PropsSI('H','T',_T,'Q',0,'SES36')

fig = plt.figure(figsize=(12,4))
ax = fig.add_subplot(121)
plt.plot(_T,_hfg,'o-',label='Riva')
plt.plot(_T,hfg,'o-',label='CoolProp')
plt.xlabel('T [K]')
plt.ylabel('$h_{fg}$ [J/kg]')
plt.legend(loc='best')

ax = fig.add_subplot(122)
plt.plot(_T,(np.array(_hfg)/np.array(hfg)-1)*100,'o-')
plt.xlabel('T [K]')
plt.ylabel('$h_{fg}$ error [%]')

plt.tight_layout()
plt.savefig('SES36_hfg.png')

[JDPS]

Thank you for the info. I will analyse this new data and contact Riva regarding this topic.