Re: [TRNSYS-users] TYPE 666 Chiller Queries

Hi Jeff,

Thanks a lot for your insightful comments to the queries raised !

I followed your lead and found the cause of the problem. The root cause actually came from an erroneous data input in the part load file being fed to the chiller.

As a result, when I increased the ECWT and consequently decreasing the capacity of the chiller, the chiller was running at a higher part load and hence steering away from the erroneous part load data that gave the unexpected spikes. When the part load data was corrected, the chiller is behaving as expected.

Now all the phenomenon observed can be explained !

Once again, I would like to thank you for pointing me to the correct direction of investigation.

And thanks to everyone who had spent their time to think through the queries raised.

Best Regards,
Yeo

From: Jeff Thornton [mailto:thornton@tess-inc.com]
Sent: Thursday, 28 March, 2013 4:19 AM
To: #YEO KEAT KEE#
Cc: TRNSYS users mailing list at the Solar Energy Lab, UW-Madison
Subject: RE: [TRNSYS-users] TYPE 666 Chiller Queries

I would run a simulation with just the chiller in it and vary one factor at a time to make sure the chiller is behaving as expected. If I was a betting man, I would say that the part load impacts are what are driving your COP swings. As the ECWT decreases, the capacity of the machine increases and for a given load, the part-load ratio decreases. If you have a rather part-load curve on the chiller it would explain what you are seeing.

Jeff

---

Jeff Thornton

President - TESS LLC

22 N. Carroll Street, Madison WI USA 53703

Office: (608) 274-2577 Fax: (608) 278-1475

www.tess-inc.com

E-Mail: thornton@tess-inc.com

On 03/27/2013 2:29 am, #YEO KEAT KEE# wrote:

Hi Jeff,

You are definitely right that there are spikes in the power plot that complements the spike in the COP.

Using the same performance map, and running at known conditions, the model is actually predicting the expected results.

The problem arises when I try to run the chiller at part load, in this case, the chiller is actually running at half of its capacity.

The spikes observed were not present when the chiller was made to run at its full capacity.

I would like to expand the discussion by bringing in the Entering Condenser Water Temperature (ECWT) VS the COP.

According to the default chiller performance map, as the ECWT is lowered, the COP ratio of the Chiller will increase, all other parameters kept constant.

In the attached file, the first picture COP VS ECWT ,shows the reverse instead, when the ECWT Is lowered, the COP of the Chiller decreases.

In the second picture, when I ramp down the fan speed of the cooling tower further, and hence increasing the ECWT, the COP of the chiller stabilized.

Does anyone knows why is such relationship being observed ? Or is there be a fundamental mistake in my understanding of the program ?

Best Regards,

Yeo

From: trnsys-users-bounces@cae.wisc.edu [mailto:trnsys-users-bounces@cae.wisc.edu] On Behalf Of Jeff Thornton
Sent: Wednesday, 27 March, 2013 11:58 AM
To: TRNSYS users mailing list at the Solar Energy Lab, UW-Madison
Subject: Re: [TRNSYS-users] TYPE 666 Chiller Queries

Well if you look carefully at your plots, power is not "almost constant". Change the scale of the power plot and you'll see the spikes that you are observing in your COP. Have you spot-checked your performance map by running the chiller at "known conditions" and seeing that you're getting out of your model what you are expecting?

Jeff
---
Jeff Thornton

President - TESS LLC

22 N. Carroll Street, Madison WI USA 53703

Office: (608) 274-2577 Fax: (608) 278-1475

www.tess-inc.com

E-Mail: thornton@tess-inc.com
On 03/26/2013 10:44 pm, #YEO KEAT KEE# wrote:

Dear Fellow Trnsys Users,

I am running a parametric study using the TYPE 666 Chiller and I have some unexplainable observations with regards to the Chiller COP.

It seems that the chiller COP has a lot of unexplainable spikes and fluctuations throughout the simulated year.

This is observed despite

1.       An almost constant part load ratio (load) handled by the chiller.

2.       Entering Condenser Water Temperature to the chiller remains almost constant, with little fluctuations.

3.       Power consumed by the chiller remaining almost constant.

I understand that in TRNSYS, COP=Qmet/P, if both the Qmet and P are relatively constant, what could be causing the spike observed in the COP ?

I have attached some screen shots from the results of my simulation.

Has anyone encountered the same situation before ? Or any advice on why this is happening ?

Best Regards,

Yeo