[esp-r] Re: unrealistic energy demands?

[Jon Hand]

A recent question about energy demands...

::I am calculating the annual heating and cooling energy consumption for office buildings in UK climates. 
::The heating and cooling setpoints are 20 and 24 degrees respectively. Internal casual gains from 
::occupants, lighting and electrical equipment are assumed. The U-values for building fabrics are 
::in accordance with the UK building regulations.

::To my surprise, the results (annual heating and cooling energy consumptions) are around 35 and 16 
::(in kWh/m2 heated floor area). The values are a bit unrealistic compared to the DETR publication for 
::energy consumption for standard air conditioned office buildings in the UK climates (128-226kWh/m2  
::for electricity and 97-178kWh/m2 for fossil fuels, or 200-260kWh/m2 for heating and 33-41kWh/m2 for cooling).

::I had tried different simulations with variable heating and cooling capacities and temperature setpoints,
::it looks to me that my office building never meet the standard guidelines established by DETR.

-----------------------

Searching for the causes for unexpected results is sometimes difficult for the author
of a model to do (many reasons for this) and ESRU always recommend that someone (else) in the
group/company is trained to QA and review models.  Often another person will spot in
a few minutes what the issue is and sometimes it is subtle and can take hours of effort.

In this case someone in the community is asking for help and it will probably require that
someone else carefully look at the model.  If the model is large. poorly documented and
the entities in the model have names that are opaque then this task will be less than 
pleasant task.  There are commercial services that will QA models.  And often users in  the
community will help out.

Prior to getting others involved it would be useful to consider whether one office of this
building could be modelled with the same assumptions as used in the bigger model
and see if the performance per m2 is also not meeting expectations.  A small model
could then be looked at by others with less hassle.

It would be useful to consider - has the simulationist generated other models which
do give expected results and if so what is different about the office building?

And to the simulationist who asked the question - if someone offers to check your
model, please first send them a QA report generated in the project manager with
all of the options toggled to verbose.  Reviewing one report is often easier than
passing a whole model across the internet.

Last year ESRU and SESG hosted a simulation QA workshop where we looked at
a number of classic omissions and typographic errors in models and how to
spot them.  Perhaps we should do another one?

-Jon Hand