I am studying the impacts of occupant behavior on the success of various building technologies, such as natural ventilation, radiant ceilings, phase-change materials, automatic lighting controls, demand-controlled ventilation, etc. using whole building energy simulation.? I am wondering whether to use EnergyPlus or ESP-r.? An advantage of ESP-r seems to be the SHOCC module, which seems to offer a great deal of flexibility for manipulating occupant behavior, including stochastic processes.? However an advantage of EnergyPlus seems to be its extensive HVAC libraries so that I can actually model energy use instead of loads.? I am wondering if the EnergyPlus Runtime Language (ERL) (described in Ellis, Torcellini, Crawley 2007) would offer the opportunity for a similar level of occupancy control via simple programming.? I would appreciate any advice on the appropriateness of each software for this research.?
Thanks in advance!
Holly
?
Holly Wasilowski
ESPr vs. E+ for occupant behavior research
Apr 9, 2010
Holly,
You can use an external file within a schedule to model say occupancy. If
you had data that told you the level of occupancy for each room for each
hour of the day, etc then you can bring that file into the EnergyPlus model
within a schedule for occupancy and hence it would allow you to model it
exactly if necessary. Then using the Parametric tools that E+ offers, you
could then change a variable at a time (that would be my preference, even
though you can do more than 1 at a time) to then work out the effects of
those variances and the link to occupancy and effectiveness, etc.
ESP-r will probably run faster than E+ which will be useful if your models
are large. Both are difficult to work in.
Cheers
Dr Paul Carey
Holly,
In EnergyPlus, you could implement a model for the user behavior using
either the Energy Management System or the Building Controls Virtual Test
Bed. The Building Controls Virtual Test Bed would allow you to implement the
user behavior model either in MATLAB, Simulink, Modelica, Ptolemy or in your
custom program, and link it during run-time to EnergyPlus so they can
exchange data as the simulation progresses. It is available from
https://gaia.lbl.gov/bcvtb
All the best,
Michael Wetter
Thanks Michael
Clearly I have a lot to learn about what E+ can link with.
Cheers
Paul
This is getting to be a good discussion. On the ESP-r side there are many options for
getting complex patterns of occupancy, lights, equipment. At one extreme you could
define for every minute of the year casual gains (occupants/lights/small power) for
one (or more) rooms in what we call a temporal file and read these in during the
simulation (or every 5 minutes or 10 minutes etc).
ESP-r has a (not-used-very-often) formal description of uncertainty so that you can
nominate portions of a model (e.g. the conductivity of brick on the south facade
or the size of a window) and say how uncertain you are about that and then direct
the simulation engine to run however many simulations are needed to explore the sensitivity
of predictions to changes in these parts of the model. There are a half dozen
formal statistical methods implemented so if you don't like a step-wise change you can use a
different approach.
What would be really nice would be to use a descriptive language that
implemented patterns of diversity overlayed onto the simplistic schedules that
most of us tend to use. I think that SHOCC does this, but it
is a bit of a learning curve for practitioners.
-Jon Hand
Holly,
If you look at the my occupancy behavior study in IBPSA2009 (I guess you
already looked), I used E+, using external input files.
At that time (early 2008), bcvtb is not mature to be implemented. However,
now I think BCVTB should be the most suitable one. It allows to interact E+
and MATLAB every timestep. I would program my occupancy algorithm in MATLAB
and interact with E+ through BCVTB.
If you have any questions on occupancy behavior, let me know.
Cheers,
Bing Dong