<br><font size=2 face="sans-serif">Dear All,</font>
<br>
<br><font size=2 face="sans-serif">Thanks for the feedbacks, they were
all very helpful. Answering back to thank and to clarify how I (hopefully)
solved the problem, which may help future users in the same situation.</font>
<br><font size=2 face="sans-serif"> </font>
<br><font size=2 face="sans-serif">Moving the components solving order,
especially placing the storage tank in the end, helped a lot to improve
the convergence but did not solve the problem completely. But it did give
me a good idea of where the main problem actually was (before I was being
misled that it was focused in the storage tank).</font>
<br>
<br><font size=2 face="sans-serif">The circuit leaving the load was the
reason for the problem on convergence. In this circuit, part of the water
returning from the load would be recirculated to keep the load feeding
temperature around a given value. The combination between this configuration
with the way the load demand is inserted (external file controlling the
flowrate) was giving me some sort of "circular reference". Because
of it, did not matter what I do, one component would always present at
some point a convergence problem. The solution was to add a type 661 component
before the load. This"opened" the loop, vanishing all the convergence
problems.</font>
<br>
<br><font size=2 face="sans-serif">Once again, thanks for all the feedbacks.</font>
<br>
<br><font size=2 face="sans-serif">Kind regards,</font>
<br>
<br><font size=2 face="sans-serif">Lucas Lira</font>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>Michaël Kummert <michael.kummert@gmail.com></b>
</font>
<p><font size=1 face="sans-serif">13/11/2009 22:24</font>
<td width=59%>
<table width=100%>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td><font size=1 face="sans-serif"><trnsys-users@cae.wisc.edu></font>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td><font size=1 face="sans-serif"><lucas.lira@scottish-southern.co.uk></font>
<tr valign=top>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td><font size=1 face="sans-serif">RE: [TRNSYS-users] Convergency problem
- using type 534</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><tt><font size=2>Lucas,<br>
<br>
You have already tried two of the “usual suggestions” to get a simulation
to<br>
converge: reducing the time step and making sure tank nodes are not too<br>
small compared to the volume of fluid that enters them during one time
step.<br>
I am afraid can only make generic recommendations, but here they are:<br>
<br>
- Try re-arranging the components in a different order (“Component order”<br>
tab of the control cards). I usually tend to place the tanks at the very
end<br>
of the calling order, the logic being that they have more thermal mass
so<br>
it’s no big deal if the other components use the value of previous time
step<br>
as a starting point. I also try to put the controllers after the components<br>
they depend on for the control decision, and for the rest I try to use
a<br>
logical order. Ooviously you have to make a choice at some point as most<br>
system have closed information loops so the order cannot be completely<br>
logical. This sounds a bit like wishful thinking but it has worked in many<br>
cases for me.<br>
<br>
- Add pipes in the model if you haven’t done so until now. Usually anything<br>
with thermal mass or time delay helps with convergence, although in your<br>
case the storage tank seems to be at the heart of the problem.<br>
<br>
- Make sure the flowrate controls are synchronized with the actual system<br>
operation. I have seen some simulations where a device was set off (e.g.
a<br>
heat pump) without turning the pump off at the same time. If that loop
is<br>
connected to, say an immersed heat exchanger in a tank, you could end up<br>
with a closed fluid loop with very small temperature differences that cause<br>
non-convergences.<br>
<br>
What usually works for me is a combination of all of the above (including<br>
what you did already) – and lots of coffee… <br>
<br>
Best wishes,<br>
Michaël<br>
<br>
____________________________<br>
Michaël Kummert<br>
École Polytechnique Montréal<br>
<br>
<br>
<br>
From: lucas.lira@scottish-southern.co.uk<br>
[mailto:lucas.lira@scottish-southern.co.uk] <br>
Sent: 13 November 2009 09:08<br>
To: trnsys-users@cae.wisc.edu<br>
Subject: [TRNSYS-users] Convergency problem - using type 534<br>
<br>
<br>
Dear All, <br>
<br>
I am currently simulating a system where 2 different energy units are<br>
connected to the same storage tanks at different nodes. In a 20 nodes tank<br>
the first will be connected between nodes 3 and 7 and the second between<br>
nodes 9 and 20. The first will be supplying water at around 55C and
the<br>
second unit at around 40C, with a higher flowrate. The load is connected<br>
between nodes 1(extracting from tank) and 18 (return). <br>
<br>
The idea of the project is to use the tank stratification to keep the second<br>
energy unit running at as low temperature as possible. <br>
The problem that i am getting is that the system is not converging.<br>
Different units are listed with this problem but the storage tank is<br>
involved in all cases, and some times the convergence problem involves
the<br>
tank alone. <br>
<br>
Trying to identify if the problem was related with control units operating<br>
in too large time steps i reduced its value down to 30s and later eliminated<br>
all control units. The convergence problem still happens specially at the<br>
storage tank. <br>
<br>
The second guess is problems regarding the heat capacity in the tank, which<br>
I am not sure how to solve. The number of nodes were reduced from the<br>
initial 20 to 10, rising each node capacity, but without success. <br>
<br>
The inversion mixing flowrate ( type 534 tank) was originally set to -1
but<br>
then changed to "amount of water in each tank/ per time step"
( In a 8000 L<br>
tank with time step of 5 min and 10 nodes, the flow was set to 9600 kg/h).<br>
Even although now inversion was being observed, type 534 was still<br>
presenting convergence problems. <br>
<br>
Rising the tank size to 80 000L the problems disappears. <br>
Using the original configuration (20 nodes 8000L capacity and no inversion):<br>
<br>
- Depending of the connection points the convergence
problem does<br>
not occur. <br>
- Depending of the first unit's rated power
(when reduced), the<br>
convergence does not occur. <br>
- Depending of
the losses through the tank (if high), the<br>
convergence problem seems to not occur. <br>
<br>
I am not sure what I can do to get the required results. I am simulating
a<br>
real case, therefore unable to change tank and energy unit sizes or flow<br>
rates. I do have a strong feeling that the problem is related with the
heat<br>
capacity through the nodes in the tank but can't find or think of a solution<br>
(even though small time steps should also make difference, shouldn't it?).
<br>
<br>
I would appreciate any idea of how to proceed or suggestion of what else
can<br>
be wrong and I may be missing. <br>
<br>
Kind regards <br>
<br>
Lucas Lira <br>
**********************************************************************<br>
The information in this e-mail is confidential and may be legally<br>
privileged. It may not represent the views of Scottish and Southern Energy<br>
Group. It is intended solely for the addressees. Access to this e-mail
by<br>
anyone else is unauthorised. If you are not the intended recipient, any<br>
disclosure, copying, distribution or any action taken or omitted to be
taken<br>
in reliance on it, is prohibited and may be unlawful. Any unauthorised<br>
recipient should advise the sender immediately of the error in transmission.<br>
Unless specifically stated otherwise, this email (or any attachments to
it)<br>
is not an offer capable of acceptance or acceptance of an offer and it
does<br>
not form part of a binding contractual agreement. Scottish Hydro Electric,<br>
Southern Electric, SWALEC, Atlantic Electric and Gas, S+S and SSE Power<br>
Distribution are trading names of the Scottish and Southern Energy Group.
<br>
Scottish and Southern Energy plc, Inveralmond House, 200 Dunkeld Road,<br>
Perth, Perthshire, PH1 3AQ. Registered in Scotland Number. 117119<br>
**********************************************************************<br>
<br>
</font></tt>
<br>
<pre>**********************************************************************
The information in this e-mail is confidential and may be legally privileged. It may not represent the views of Scottish and Southern Energy Group. It is intended solely for the addressees. Access to this e-mail by anyone else is unauthorised. If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful. Any unauthorised recipient should advise the sender immediately of the error in transmission. Unless specifically stated otherwise, this email (or any attachments to it) is not an offer capable of acceptance or acceptance of an offer and it does not form part of a binding contractual agreement. Scottish Hydro Electric, Southern Electric, SWALEC, Atlantic Electric and Gas, S+S and SSE Power Distribution are trading names of the Scottish and Southern Energy Group.
Scottish and Southern Energy plc, Inveralmond House, 200 Dunkeld Road, Perth, Perthshire, PH1 3AQ. Registered in Scotland Number. 117119
**********************************************************************
</pre>