Changing timestep effects simulation results highly? Insulation and window U values very low impact.

Hi. I am modelling a very inefficient library building in E+ which has a CAV system and very low insulation etc. Whenever I change the timestep to a different number, the results change drastically. For example: for the values of timesteps 1, 5, 20, 30; the corresponding results are 406, 322.5, 265, 249.13 kWh/m2y.

Another weird behavior with the model is that when I change the values for insulation thicknesses for the exterior walls or the U values of WindowMaterial:SimpleGlazingSystem, the results are almost not effected. (I have checked that the materials I produce are referenced in a construction object and that construction object is also referenced inside all the exterior wall surfaces of the building!). When the insulation (polyurethane) thicknesses are 0, 3 cm and 5 cm; the corresponding results are 277.02, 265.79, 263.37 kWh/m2y. When the window U-Values are 5.8, 3.0, 1.9; the corresponding results are 277.02, 268.04, 265 kWh/m2y.

I feel like the timesteps should not effect the simulation results this much and changing the window U values and insulation thicknesses should change the results much more than how much it is effected here. Could this be showing a huge problem with the way that I modeled the building or could this be a normal behavior that should be expected?

Any help is appreciated. (I couldn't upload my idf file due to my karma level. ) (I am using Energyplus version 9.3)

ofkurt

asked 2022-08-13 08:09:13 -0500, updated 2022-08-17 04:41:00 -0500

edit flag offensive 0 remove flag close merge delete

Comments

add a comment see more comments

1 Answer

Sort by

oldest

newest

most voted

After many trial simulations, I was able to find the problem in my model. It was caused by having 2 setpoint managers on the same node and the minimum part load ratios of the boilers was set to 1. After realizing that this was the case, I came across to a warning in the Input Output Reference Document related with having two setpoint managers on the same node. I didn't know this when I was producing the model. And because I was trying so many different things, I think I left the minimum part load ratio as 1, not knowing how it would effect the simulations.

(I am still working on finding out why the insulation thicknesses don't effect the simulation results as expected..)

From Input Output Reference Document (1.49 Group Setpoint Managers):

No two setpoint managers should use the same setpoint node for like control types. For the case when two setpoint managers place a setpoint on the same setpoint node (e.g., a temperature setpoint placed on the same node by two diﬀerent setpoint managers), the setpoint value on the node is calculated by the last setpoint manager simulated. Within a specific type of setpoint manager (e.g., Scheduled), the setpoint managers are simulated in the order found in the input data file (idf) when viewed using a text editor. It is also possible for two diﬀerent types of setpoint managers to place setpoints of like control variable on the same node. In this case, the order of simulation for the type of setpoint manager determines which setpoint manager is the last to write a setpoint value to the node. This conﬂict in setpoint node data is most likely to cause inaccurate simulation results. For this reason, a conﬂicting setpoint node name warning message is issued so that the conﬂict can be resolved. Duplicate setpoint node names found within a specific node list will not typically cause problems with the simulation (the same value is written twice to the same node). In this case a duplicate setpoint node name warning is issued, however, this could be an error in the spelling of the node name and should be corrected.

ofkurt

answered 2022-08-17 04:54:39 -0500