Multiple year RunPeriod tabular results

[JasonGlazer]

When running a multiple year simulation for ground loop heat exchangers (see RunPeriod field named "Number of Times Runperiod to be Repeated") the tabular reports show the totals over the entire run period. This probably does not make sense since the idea of running multiple years is because the ground needs to reach a steady annual cycle. Instead only the last year of the simulation when multiple years are requested should be reported. It might make sense to make this an option so that reporting could be for only the last year or for all years cumulatively. Any test file when changing the RunPeriod field named "Number of Times Runperiod to be Repeated" to a value 2 or greater should show this problem.

[JasonGlazer]

@mjwitte @Myoldmopar Before I fix this issue we need to decide what we want as the solution. Currently the multiple years are just added together to show the sum of the results in the reports. This does not seem correct for a ground loop heat exchanger but would probably be ok for a custom weather file that is multiple years, say of actual weather data. Someone who is using EnergyPlus for simulating multiple actual weather years would probably not want us to change the current behavior. In the multiple year weather file, what would really be good is having some of the tabular reports repeated for each year and then a summary combining the multiple years. I guess I am coming to the conclusion that their should be some control, either in the RunPeriod:* objects or perhaps a new OutputControl type of object that lets the user select the option they want. I think the options would be ReportLastYearOnly, CombineMultipleYears, ReportEachYear, and ReportEachYearWithCombinedSummary. If we are going this route, I think maybe this should be considered for a future enhancement rather than an Issue. What do you think?

[mjwitte]

@Myoldmopar @JasonGlazer I would just change things to report the last year for now. If someone wants to see some intermediate year results they could use output:variable/meter to see that. Or can they? What happens, say if you report an output:meter at the "environment" frequency? Does it report at the end of each repeated year, or just the cumulative at the end of the last year?

[Myoldmopar]

I don't know off the top of my head, but I would hope that "Annual" would report for each yet year, and "Environment" would report once. For the "RunPeriod" frequency, I think the it should be reported at the end of each year also, since the multi-year is generated from the IDD field Number of Times Runperiod to be Repeated, thus implying each run period is one year, for example.

[JasonGlazer]

According to the IDD, RunPeriod, Environment, and Annual are synonymous.

I verified this,using any of these just results in a single result at the end of the simulation.

[JasonGlazer]

@mjwitte In testing the routine that zeros the accumulation arrays, I have noticed that the same file run for one year and running for three years and only using the last year of results, give similar but not exactly the same results in the tabular reports. Is that reasonable or should I expect the results to be exactly the same? I am using 5ZoneAirCooled.idf as the basis of my testing.

[mjwitte]

I would expect a small difference in the first week of January due to different wall history in the first year (repeated Jan 1) vs the later years where it just continues from Dec 31. I would expect years 2 and up to agree with each other exactly.

[JasonGlazer]

I think the day of the week also changes since Jan 1 follows the day of week from Dec 31. I am comparing only the last year with a simulation lasting only one year and finding differences. Not sure what should be considered reasonable differences.

[JasonGlazer]

@mjwitte @Myoldmopar The tabular report called "Energy Meters" is very closely tied to the energy meters themselves. I cannot reset to zero the variables used without affecting the results that are reported in the MTR file which, when requested, are for the entire runperiod. I have not been touching the MTR file or the ESO file. I could make some additional elements in the MeterType to accumulate and do max and min for the last year of the simulation only but it seems to be adding code that is very redundant. Another instance of lines in UpdateMeter(). So I want your opinions, do I

a) leave the Energy Meters tabular report alone and do not show only the last year?

b) change the results in the MTR file when someone wants a result for a runperiod that is more than a year so that it reflects only the last year?

c) add new elements to the MeterType and new lines to UpdateMeter() to have a new set of accumulating values that are only for the last year of a multiple year runperiod?

After laying these options out, I think I am going to go with the (c) option after all. It accomplishes the goal of fixing the issue and will not impact the MTR reporting. Please let me know if you disagree.

[mjwitte]

@JasonGlazer Regarding the day of week, there is a field in RunPeriod that controls this "Increment Day of Week on repeat". Setting that to No may remove some of the differences.

My expectation for repeated runperiod would be that the mtr and eso files would report the first repetition, reset, then report the second repetition, etc. Regardless of the reporting frequency.

What happens currently when you have multiple runperiods (say, Jan, Apr, Jul)? Do the max/min values reset?

More flexibility may be required later when multi-year (not repeating, just longer) runperiods support is added. Regardless, the I/O Ref and IDD notes for Runperiod "Number of Times Runperiod to be Repeated" should explain what will happen with reporting.

[JasonGlazer]

@mjwitte I changed the "Increment Day of Week on repeat" to No and the magnitude of the differences is much smaller, thanks. This is issue is specifically about the tabular reports and not the MTR and ESO files, if we want to change their behavior, maybe we need to post another issue.

[mjwitte]

But all outputs need to be consistent with each other. The only thing that would change in the MTR and ESO files is the max/min values, correct?

[Myoldmopar]

Closed with #5256