Re: [TRNSYS-users] Simplified heating and cooling based on operativetemperature

Dear all,

I also tried the approach suggested by mr Olsen, but a little different since I don't know how to change the deck. What I did was (see also attached file):

- I set the setpoint temperature in TRNbuild as an input. (see figure 1)
- Then I calculated the adapted air setpoint temperature (Tair_set) in TRNSYS studio as follows: (fig 2)
Tair_set = (Tset_WK-0.5*Tms)*2 , where:
- Tset_WK is the required operative temperature for a weekly schedule
- Tms is the CURRENT mean surface temperature of the zone, which is a standard output of type 56. (outp Ntype24)

This way the current mean surface temperature is used. To me the results seem OK (see also attached file). Does anyone have an idea if this is a correct approach or that something might be wrong with it? (in my case the difference in temperature of Tmeansurf is up to 0.4 K, since I use night setback I think.)

Thank you a lot!

If I can be of any help to you, let me know.

Kind regards, Sabine Jansen

p.s. I have 2 more questions about the ideal heating demand + one general question. I will also ask these to the general users list, but since they are related I mention them here.

1.    it takes two timesteps for the ideal heating to reach the required setpoint..
Is this normal or what am I doing wrong?

2.   When using a schedule in trnbuild (so not in the operative correction trick of above), somehow the heating only switches on at exactly the time of the switch (in my case 7 h and 17 h), while I assumed the schedule to mean that at the new time the new temperature must be obtained (see attached document). The heating switches OFF at exactly the changing time, so it then takes the new setpoint temperature. (see figure 2).

3.   Does anyone know why the temperature seems to be fluctuating a little (in other words: it is not a smooth line) as soon as the heating turns off?

Also when using a smaller timestep the line is not smooth, but the temperatures stay constant during a hour, so the line looks like a staircase..

     (note: the non-smooth line occurs when the heating is OFF)

Van: Maarten Sourbron [mailto:Maarten.Sourbron@mech.kuleuven.be]
Verzonden: vr 19-2-2010 11:55
Aan: trnsys-users@cae.wisc.edu
CC: Kersken, Matthias; Sabine Jansen - BK; olsen@transsolar.com; keilholz, werner
Onderwerp: Re: AW: [TRNSYS-users] Simplified heating and cooling based onoperativetemperature

Sabine Jansen

PhD Candidate

www.lowex.nl

Faculty of Architecture, TU Delft

Room : 01+WEST220
Julianalaan 134
2628BL Delft

present on Monday, Wednesday & Friday

T: +31 (0)6-42428622

F: +31 (0)15 27 84178

E: S.C.Jansen@tudelft.nl

Sabine Jansen

PhD Candidate

www.lowex.nl

Faculty of Architecture, TU Delft

Room : 01+WEST220
Julianalaan 134
2628BL Delft

present on Monday, Wednesday & Friday

T: +31 (0)6-42428622

F: +31 (0)15 27 84178

E: S.C.Jansen@tudelft.nl

Thank you all for your answers : I checked the different solutions proposed.

With the type22 (iterative feedback controller) it seems to work but for some reason the controller seems satisfied with a resulting temperature lying outside the imposed tolerance (tol = 0.2K while Toperative - Tset = 0.8K sometimes). I didn't check the controllers algorithme to find out what's going on, though. The controller status indicated for most of the timesteps a value of 65 (1 + 64), which means the u-e - curve is 'too steep' for the controller. This might cause the problem?

With the reverse procedure suggested by mr. Olsen, it works fine, except from the fact that the mean radiant temp used to calculate the 'corrected' T_air, is always from the previous timestep (or am I mistaking here?). But, since Trad_mean doesn't change quickly, this only leeds to a minor deviation (0.1K max).

The Top-equation suggested by mr. Keilholz is more detailed, but the PID controller I tried quickly didn't manage to keep Top constant. I didn't look at it very carefully to find good controller parameters, I admit.

Thank you all!

Kind regards,

Maarten

Kersken, Matthias schreef:

Hello Ms. Jansen and Mr. Sourbron!

I also think the solution has to be outside Type56.

If You use an iterative-feedback-controller (Type22) instead of Type2 and feed the operative temperature in as "controlled variable", define Your Setpoint and give the output of the controller as Gain in the thermal zone You should be able to avoid the hysteresis-effect occoring when using Type2. (You can also set a maximum output/heating power).

Using Type2 a bigger Deadband or a shorter timestep my also solve the problem.

With kind regards,

Matthias Kersken

---

Dipl.-Ing. Matthias Kersken
Wissenschaftlicher Mitarbeiter / Abteilung Energiesysteme

Fraunhofer-Institut f�r Bauphysik
Institutsteil Holzkirchen

Fraunhoferstra�e 10
83626 Valley

eMail: Matthias.Kersken@ibp.fraunhofer.de
Telefon 08024 / 643-204
Telefax 08024 / 643-366
www.ibp.fraunhofer.de

Dipl.-Ing. Matthias Kersken
Research Associate / Department Energy Systems

Fraunhofer Institute for Building Physics
Branch Holzkirchen

Fraunhoferstra�e 10
D-83626 Valley

eMail: Matthias.Kersken@ibp.fraunhofer.de
Phone +49 8024 / 643-204
Fax +49 8024 / 643-366
www.ibp.fraunhofer.de

---

Von: Sabine Jansen - BK [mailto:S.C.Jansen@tudelft.nl]
Gesendet: Donnerstag, 18. Februar 2010 14:51
An: Maarten Sourbron; trnsys-users@cae.wisc.edu
Betreff: Re: [TRNSYS-users] Simplified heating and cooling based onoperativetemperature

Dear all,

I am also interensted in that question.

The only way I know now how to solve this is by using a type (controller, type 2 or thermostat, type 8) outside type 56. The the operative zone temperature must be connected to this type (e.g. 2b and creating a schedule elsewhere, in an equation solver), and the heating must be defined as a limited power connected to an input, which is then the output of the control type (type 2 or type 8).

One problem with this approach is that I cannot use a very large heating power anymore, since that causes the controller to switch on and off. Thus the heating power must be determined using the ideal heating, based on air temperature, first.

If there is an easier way, I am also very happy.

Kind regards,

Sabine Jansen

Van: Maarten Sourbron [mailto:Maarten.Sourbron@mech.kuleuven.be]
Verzonden: do 18-2-2010 13:53
Aan: trnsys-users@cae.wisc.edu
Onderwerp: [TRNSYS-users] Simplified heating and cooling based on operativetemperature

Dear all,

The simplified heating and cooling in Trnbuild is now comparing the zone
AIR temperature with the given set points. Is there a way to change
this (or another controller model) to have the zone OPERATIVE
temperature compared with the set points?

Kind regards,

Maarten Sourbron

--
Departement Werktuigkunde-Department of Mechanical Engineering
Katholieke Universiteit Leuven
Celestijnenlaan 300A bus 2421
3001 Heverlee (Leuven)
T : 0032 (0)16 32 28 73
F : 0032 (0)16 32 29 85
E : Maarten.Sourbron@mech.kuleuven.be

_______________________________________________
TRNSYS-users mailing list
TRNSYS-users@cae.wisc.edu
https://www-old.cae.wisc.edu/mailman/listinfo/trnsys-users

-- 
Departement Werktuigkunde-Department of Mechanical Engineering 
Katholieke Universiteit Leuven
Celestijnenlaan 300A bus 2421
3001 Heverlee (Leuven)
T : 0032 (0)16 32 28 73
F : 0032 (0)16 32 29 85
E : Maarten.Sourbron@mech.kuleuven.be