setpoints and setback in cooling design and simulations

mbwaite · Posts: 44 Location: New York, NY

For a summer design day, I have the cooling schedule set to "On." My cooling setpoint is 72F, but temperatures get to be above that. My setback is at 82.4, but that shouldn't matter, since the setpoint should be the control at all times.

Another problem I am now seeing is probably related. Even when I have a Zero on my schedule for a simulation, DB is still using the setback temperature to switch on the cooling at those times for which I have set that cooling is not available at all.

There seems to be a problem with the program.

mbwaite · Posts: 44 Location: New York, NY

Any help on this????

Andy Tindale · Posts: 1196 Location: Stroud

Taking the second problem first, I would guess that you are using 7/12 schedules? If so, there is a fix for this in version 1.3.4.024. Previous versions were interpreting '0' as 'use setback' instead of 'off'.

As for the first problem, could you please send the model, stating which zone is being inadequately cooled?

Thanks,

Andy

mbwaite · Posts: 44 Location: New York, NY

Andy,

This is the same very large model that we have discussed before, so it is too big to post. As for which zone it is happening in, it is all of them (695 zones). The temperature I am referring to is the "Max Temperature." I understand this is the mean of the air temperature and radiant temperature, but for some zones it reaches 85F, meaning a radiant temperature of 98F, which I find highly unrealistic. To reiterate, the cooling is set to "On," so the setpoint temperature of 72F should be the maximum air temperature.

Thanks for the help,

Mike

Andy Tindale · Posts: 1196 Location: Stroud

Mike,

Just thought I'd check before running long simulations, that you are looking at hourly temperatures within DesignBuilder?

The fact that you mentioned 'Max temperatures' makes me wonder whether you are perhaps looking at maximum temperatures over a day/month as reported in the EnergyPlus output files. The maximum values would include unoccupied periods where the cooling system may not be operational.

Andy

mbwaite · Posts: 44 Location: New York, NY

Andy,

This is for "Cooling Design" (One Day and NOT a "Simulation") and the cooling is ALWAYS "On." It is always available. At all times, regardless of occupancy.

The "Max Temperature" is an output in DB under the "Summary" tab for the "Cooling Design" Results. It lists the "Max Temperature" in each zone. The "Max Temperature" is the average of the "Air Temperature" and the "Radiant Temperature." It is not hourly data I am looking at, it is the the "Summary" for "Cooling Design" results in DesignBuilder.

Everything in quotes here is DesignBuilder's own language.

Thanks,

Mike

Andy Tindale · Posts: 1196 Location: Stroud

OK thanks for the clarification, I'll take a look.

Andy

Andy Tindale · Posts: 1196 Location: Stroud

Mike,

I think the reason you are getting temperatures higher than you expect is that the zones are being controlled to 'air temperature' but you are viewing 'operative temperatures'. If you change to operative temperature control by changing the setting on the cooling design calculation options dialog you will see temperature responses like that below with operative temperatures constant at the setpoint over the day. I have attached the equivalent temperature response for air temperature control.

Regards

Andy

mbwaite · Posts: 44 Location: New York, NY

Thanks Andy. I just thought it seemed really high that the radiant temperature in some zones got up to 36.5 C, but maybe it's right.

mbwaite · Posts: 44 Location: New York, NY

Now, every time I try to run a simulation, the program closes at some point during the middle of it. I am trying to repeat the same simulations I ran before, but with the new version, so as to avoid the problem with the setback temperatures.

I do not know at what point the problem occurs, because I will try to have the simulation run overnight, since it takes so long, but there are no warning messages or anything. The program simply closes. This has happened the last 4 times I have tried to run a simulation. This makes it incredibly difficult to use the program, as I have to do the modeling and have deadlines. While it is understandable that the simulations will take a long time to run, this problem is unacceptable.

Andy Tindale · Posts: 1196 Location: Stroud

Mike,

Is it possible that the simulation is crashing because you are generating hourly data for an annual simulation and have either filled your hard disk or run out at memory when generating building and block data after the simulation? For a building having 695 zones generating and processing hourly data for the whole year would be impractical.

If this does not explain it, could you please e-mail me your file so that I can try to reproduce the problem?

Andy

mbwaite · Posts: 44 Location: New York, NY

Andy,

Thank you for your prompt response.

I do not believe it is an issue with the amount of data being generated. I am running a two-week simulation and collecting hourly weather data. In the past, when trying to run an annual simulation, I received an error message when I had maxed out my memory. Since then, I have run numerous two-week simulations and did not have a problem, aside from the issue with the setback temperature being used even when cooling should not have been available. Upon your suggestion, I downloaded the new beta version, but have not been able to get a simulation to complete with it.

I will email the file to the supportdesk address, but I imagine there may be something different with the version of DB I am using now compared to the previous version that may be causing the problem.

Thanks,

Mike

Andy Tindale · Posts: 1196 Location: Stroud

Mike,

We are still investigating the problems with your model, but here is an interim report.

I don't think the problem is related to the new version of DesignBuilder because I have managed to successfully simulate one of your previous cityline models using it. The most likely explanation is, as before, too many report variables being requested. DesignBuilder 1.3.4 automatically requests monthly results and you have also requested daily and hourly results. This amounts to an awful lot of data when you consider all of the surfaces and the zones. I am currently simulating the model without monthly and daily results and will let you know if/when it finishes.

Some other thoughts:

1. You have a block at the bottom of your building which appears to be used to model a slab. I may have misunderstood the purpose of this block, but think you may be able to dispense with it. Slabs between the building and the ground or between floors are automatically modelled simply by defining the floor slab on the construction tab.

2. The model is using steps of one hour. This may be OK, but I would recommend using time steps of 2 per hour (30 minutes). Previous experience suggests that using a longer time step does not necessarily speed simulations because each time step takes longer to converge. Also you get less resolution with changing values of schedules, solar position etc.

3. The latest version of your model you sent by e-mail uses 7/12 schedules instead of Compact schedules used in the previous version. DesignBuilder does not generate such efficient schedules for EnergyPlus with 7/12 schedules which doesn't normally matter, but in your model it may possibly be having an influence on the simulation time - I will let you know when I have more information on this.

4. I have made a change to DesignBuilder to allow you to switch on/off monthly/annual result generation. Generally users will want to include monthly results but for large models like this when you may only need hourly results it should speed the initialisation process.

Regards,

Andy

mbwaite · Posts: 44 Location: New York, NY

Andy,

Thank you very much for the help.

The thing that is most confusing to me is that I ran a bunch of models before with the same schedule and the same collected results and did not have this problem.

Regarding your other comments:

1. I know the way I have the block at the bottom is far from ideal, but I needed to stretch the lowest block down to avoid this. This produced one of two problems: either the program would crash while trying to move the bottom face downward or it would work, but all of my glazing (individually drawn) would disappear. Obviously, on my next model, I will take this into account.

2. Interesting advice regarding time steps. I had just assumed I was helping things by doing that. So, what you are suggesting is to increase the number of time steps per hour, but still report just hourly results? I will try that.

3. I did not know this would cause trouble. That schedule is pretty important though as it is intergral to understanding what is going on in some of the units in this building.

4. Thank you. I do not use the monthly results at all. I will try running it with the new version. It is easy enough to do any post-processing along these lines on my own anyway using Excel, etc.

I appreciate your and everyone else's help on this. While this analysis is a project for a client of ours that needs real results, it is also a significant learning experience for me. I have done smaller models with DB and e+, but nothing approaching this size or complexity.

Thank you,

Mike

Andy Tindale · Posts: 1196 Location: Stroud

Mike,

I forgot to mention that the new version with option to switch off monthly results is not available yet but will be in the next day or two.

Andy

Login