I have found a serious bug in shadow modelling with EnergyPlus when variable shadow is considered. Take an example as simple as the ASHRAE BESTEST 610, a low-mass building loads test, same as ASHRAE BESTEST 600 but with shade overhang on south wall above windows. I have downloaded the idf's to run BESTEST 600 and BESTEST 610 with E+ V22.1.0 from besttest-gsr.s3-us-west-2.amazonaws.com. The epw file USA_CO_Golden-NREL.724666_TMY3.epw, provided with the E+ V22.1.0 installation. I obtained these results: Zone Air System Sensible Heating Energy: H600=1.6477e+10 J/year for BESTEST 600, H610=1.6706e+10 J/year for BESTEST 600; Zone Air System Sensible Cooling Energy: C600=1.9094e+10 J/year for BESTEST 600, C610=1.3305e+10 J/year for BESTEST 610. These results are expected effects of shadowing; so, up to now, E+ seems to work well.
Now, let us make a simple modification to BESTEST 610 to allow shadowing only between March 21 and September 20 (spring and summer), giving rise to what I called BESTEST 610V. The corresponding transmitance schedule is:
Schedule:Day:Interval, Day Sch 1, Fraction, No, 01:00, 1, 02:00, 1, ... , 24:00, 1; Schedule:Week:Daily, Week Sch 1, Day Sch 1, Day Sch 1, ..., Day Sch 1; Schedule:Year, Sch 1, Fraction, Week Sch 1, 1, 1, 3, 20, Week Sch 2, 3, 21, 9, 20, Week Sch 1, 9, 21, 12, 31; !Week Sch 2 is that defined for BESTEST 610 with transmitance 0.
After running E+ V22.1.0 for BESTEST 610V, we have: Zone Air System Sensible Heating Energy: H610V = 1.6334e+10 J/year, Zone Air System Sensible Cooling Energy: C610V = 2.0250e+10 J/year.
Now, the suspicion arises: Why C610V is higher than C600?, Why H610V is lower than H600?. Then, suspicion becomes a painful truth when you realize that the hourly values of Zone Air System Sensible Heating Energy and Zone Air System Sensible Heating Energy before March 21 for BESTEST610V (with shadow only between March 21 and September 20) and those for BESTEST 600 (without shadow) are completely different, while they should be identical.
Is there a way of circumventing this so serious bug at the E+ user level, i.e., without fixing the E+ source code?
Prof. Dr. Victor Fachinotti https://www.scopus.com/authid/detail.uri?authorId=6602460999
