[TRNSYS-users] Type60t and convergence problems

[John Bisgrove]

Matteo,

 

I use type 534 and have convergence issues. The first step I use when
debugging tank convergence is to set the tank to instantly correct thermal
inversions.

Another technique is to set your pump to speed up slowly so that the system
does not see the thermal shock. Both of these limit the model’s validity,
but each is valid for debugging purposes. Type60 has a destratification
conductivity parameter. Sounds like it might possibly help you debug. Also
increasing iterations will sometimes allow the model to find a solution
although in cases of high relative flow rates or large temperature
difference the thermal shock might be too large to allow convergence at all.

Solar can be especially problematic due to the temperatures of air-exposed
and ground-exposed liquid that enters the tank during winter startup
conditions. Adding pipe to the system can help, but sometimes it makes
matters worse.

 

I hope some of this advise is useful.

Good Luck,   

 

John Bisgrove

Earth Sensitive Solutions

7523 County House Road

Auburn, NY 13021

(315)253-3779

  _____  

From: Matteo DF [mailto:deframat at gmail.com] 
Sent: Sunday, January 17, 2010 4:13 AM
To: trnsys-users at cae.wisc.edu
Subject: Re: [TRNSYS-users] Type60t and convergence problems

 

Thanks for the replies...

My timestep is 15 minutes, I need to simulate 1 to 10 years so if i use a
littler time step the calculation time raises too much... I'll try
re-ordering the components and make sure there are no very small flowrates,
in fact the flowrate out from the heat pump is calculated from the thermal
request of the building which varies every hour. In fact if this thermal
request is approaching to zero the flowrate becomes very small.

I'll check if my university has bought the TESS libraries which contains
type534...

Thanks again,
Matteo.

2010/1/15 Michaël Kummert <michael.kummert at gmail.com>

Matteo,

Sometimes re-ordering the components helps (“component order” tab in the
control cards). Since you have already reduced the time step and added pipes
to the system that’s the only thing I can really think of. You could also
check the flowrate / heat transfer rate out of the heat pump, see if they
make sense. If you get extreme values they might be causing problems in the
heat exchanger model within the tank. Sometimes very small  (but non-zero)
flowrates also cause problems.
This being said, Type 60 is one of these very old components that have not
really aged gracefully in my opinion, and I avoid using it. Matt probably
didn’t want to appear as a salesman for the TESS libraries, but I would
recommend that you give a try to the tank model (Type534) in the TESS
libraries if you can get access to them. It is easier to use and far more
robust than Type 60. Of course you can still have convergence problems in
the TRNSYS solver with that component, but at least you can trust the
results when TRNSYS converges...
Good luck,

Michaël

____________________________
Michaël Kummert
École Polytechnique Montréal



From: Matteo DF [mailto:deframat at gmail.com]
Sent: January 14, 2010 06:06
To: trnsys-users at cae.wisc.edu

Subject: [TRNSYS-users] Type60t and convergence problems

Hi everybody,

I'm developing a model which simulates a solar assisted ground source heat
pump. In this model I'm using 3 type60t tanks, one for domestic hot water,
one for the ground heat exchangers (just before the heat pump) and another
one is used for building heating.

The problem is this last tank, in fact it contains 3 internal heat
exchangers, one for heat coming from the hp, one from solar collectors and
the last one for the auxiliary heating... The problem is that if I connect
the HP to it I get convergence errors (which are not present if the hp isn't
connected...). I checked every parameter, I've modified the tollerance for
convergence from 0.001 to 0.01 and placed some type31 tubes in every section
which convoys water to the tank... But the problem persists...

How do I solve these problems? Isn't there a simpler tank which I can use
that have the possibility to include at least 3 heat exchangers?

Thanks!
Matteo.

--
La scoperta è vedere ciò che tutti hanno sempre visto e
pensare ciò che nessuno ha mai pensato.
(Albert Szent-Gyorgyi, scopritore della vitamina C e premio Nobel)




-- 
La scoperta è vedere ciò che tutti hanno sempre visto e
pensare ciò che nessuno ha mai pensato.
(Albert Szent-Gyorgyi, scopritore della vitamina C e premio Nobel)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/trnsys-users-onebuilding.org/attachments/20100118/41add806/attachment-0001.htm>