Re: RE: Voodoo Engineering

18 posts / 0 new
Last post

Since my response has ended up on Bldg-Sim, I might as well include the attachment with the response which gave my views

Varkie Thomas's picture
Joined: 2011-09-30
Reputation: 0

Great analysis of energy modelling and the short comings of various
simulation packages. I especially agree with your analysis of
equest. I would prefer a much simpler interface with the option to
edit the .inp file directly then run from the command line. One can
do this with equest but you also have to download the standalone
version of DOE2.2 and mess around with the libraries before you can
run an equest model from the command line.

Chris Jones

Chris Jones's picture
Joined: 2011-09-30
Reputation: 0

Varkie/Chris and other Bldg-Simians:

This discussion, for better or worse (it seems unintentionally), has
found its way to both the [equest-users] and [bldg-sim] mailing lists,
and seems to be progressing as distinct conversation threads... While
[bldg-sim]/[bldg-rate] seem to be the best fit considering the topic,
I'd encourage anyone interested in following/joining the conversation to
reply to the thread on [equest-users] as (1) it's evident the attempt by
some to cross-post is being at least partially blocked and (2) there
seems to be a larger active discussion there, so your thoughts posted
here will not be missed.

That's my third cent (the other two are already in equest-users)!


Nick-Caton's picture
Joined: 2011-09-30
Reputation: 805


Vast subject. Kudos for condensing it whilst conveying all the
necessary meaning.

We are now at a point where Energy Modellers are at the very least
specialist engineers. In fact, you could say the best are indeed


Chris Yates's picture
Joined: 2011-10-02
Reputation: 0

Just to add a couple of points to this interesting debate.

I see the problem being that, as we increasingly set more defined limits
regarding energy modelling and its role in building regulation, we are
seeing buildings that are being built and designed to purely meet
compliance. This is in some part is useful as it brings all buildings up to
a minimum standard, the flip side of that problem is that it also means,
that to many developers this means there is no incentive to strive for
alternatives or innovative solutions. It can also allow therefore lead to
the use of simpler tools that meet those prescribed limits, but really don't
push the boundaries of engineering design enhancement of buildings.

The correct implementation and use of energy modelling need not be a
hindrance to projects nor be seen as a necessary "extra" or evil if you
consider the design process as a whole. If you use the tools at the concept
or schematic design phases, this can quantitatively confirm an engineer's
instinct or gained experience in way that will enable them to show
compliance later on. It will then allow the team to come to a decision on
the most energy efficient but also compliant route of design earlier on in
the design stage and should stop the repeat iteration of designs as the
building design progresses and therefore reduce design costs and with luck
increase productivity and profit accordingly. Fanciful dream perhaps, but
it does work.

I visited an architect a while back and he said to me "Why do I need to do
modelling, I know the principles of good low energy design, I can read books
and learn more if I need to". To which I replied, "Well every time you
send me a job to check for building regulation compliance 3 weeks before it
goes before a planning team, I normally have to tell you what you need to do
in terms of meeting compliance as your buildings are consistently failing
and you then have to rush to make those changes. I am effectively designing
your buildings for you, so if you want to continue without using energy
modelling then please carry on, and I'll continue to design your buildings."
As you can imagine this was one of those Eureka moments for this Architect,
as I waved my red rag in front of his face.

My tuppence worth.


Paul Carey's picture
Joined: 2011-10-02
Reputation: 0

I do not do modeling on a daily basis so I'm not as experience as many
other here. I do agree modeling just for LEED is silly. I have seen
modeling inform designs, reduce loads, and SIZE SYSTEMS. This last part
is what most bugs me. Why do people put so much effort into models and
then not use them to size the systems? Preventing over sizing is a great
benefit of modeling. What is your experience with using models to size
systems? Why do engineers fall back on the vendor based programs and 9
out of 10 times end up over sizing systems?

Brad Acker, P.E.

Brad Acker's picture
Joined: 2011-10-01
Reputation: 0


Using energy modeling for system sizing would be a misuse of a tool.

The energy model uses average temperatures, design uses extreme

Another example would be a classroom which can hold up to 40 people.
Actual use is estimated at 25 people, but owner would like to be able to have
40 people. The model would use 25 people (actual use) but the designer would
size equipment for 40 people (worst case).

Also, energy models get to count rejected heat from office equipment
and people. When sizing equipment you can not count the lighting, office
equipment and people heat to assist in heating.

There is more CYA in equipment sizing. There is more liability.

Jeurek's picture
Joined: 2010-10-07
Reputation: 0


Using energy modeling programs for sizing is another topic that we'll find a
lot of people on the list disagreeing about.

Really, Trane Trace and Carrier HAP are equipment sizing programs that also
do energy modeling.

eQuest/DOE 2.2 (the program I am most familiar with) also has the option of
setting design day schedules, and will give you reports that show peak
Heating and Cooling loads. With sound engineering judgment, a thorough
understanding of the modeling program you're using, and careful checking of
your results, I believe that you can use many hourly modeling programs to
size your equipment, or at least to provide a "second opinion" about the
results from your primary sizing program.


Steven Savich's picture
Joined: 2011-09-30
Reputation: 200
cmg750's picture
Joined: 2010-10-05
Reputation: 0

So let me ask you John,

How do you find peak heating and cooling loads without using a modeling program such as HAP or Trane Trace to take into account equipment loads and solar heating loads? Of course you can't use equipment loads to find peak heating load, but you can use them to find peak cooling loads. Do you only size heating equipment? And don't you realize that modeling software can be changed for different occupation conditions? If you are using effective modeling software, there are a multitude of factors and situations that are taken into account. Its not the software itself that is effective, it is the person who is using it intelligently ; )


Matthew Kimball)'s picture
Joined: 2011-10-02
Reputation: 0


Two points:

1. What Steven said; use equipment sizing software to do the energy model.
Scheduling & weather are the only differences.

2. Energy models will not go away, what will go away is the willingness to
accept someone's word. We need energy models to support design decisions
to ourselves and others.


Bonafe, Wes's picture
Joined: 2011-10-02
Reputation: 0


I use trace for Sizing and energy modeling. But I have 3 different files.

1. Equipment sizing - This has the max number of people ever expected in the
building. So if it was a school I don't count a child in the classroom and
the lunch room (he is either in one place or the other). I set the lights
and people to be cooling only. This is used to size chillers, boilers, the
other plants.

2. System sizing - I call this a butt in every seat. This has the max
number of people in each zone. I use this to size VAV boxes and other zone
equipment. The number of people is larger than reality.

3. Energy Model - This has schedule, the office equipment schedules (heat
rejection counts), I input data about equipment performance, drift points....
Well you know about energy modeling.

What you can't do is print out your enegy model and try to use the results to
size equipment. (from what I understand)

I use the same program, same construction, same temperature setting but each
file has a different intent. I change the weather from 99.5% to the 8760, I
change the number of people from worst case, to actual.

There many be other methods, but this is how I do it.

Jeurek's picture
Joined: 2010-10-07
Reputation: 0


I did HVAC design consulting for 14 years, and I used to oversize
systems just like everyone else. Now I am an energy conservation
programs evaluator, and have discovered the extent and negative impacts
of over sizing. Not speaking for all designers who oversize, I did it
mainly out of willful ignorance. It was easier for me to loosely
calculate the peak loads and then beef those estimates up enough to
safely cover and mistakes or false assumptions than it was to calculate
the loads with enough confidence to properly size the systems. I
believed in the old adage that too much is just right from my
perspective, because it minimized risk.

But now I have to measure the effects of excessive over sizing and see
the results, some of which are poor humidity control, short cycling,
reduced system operating efficiency and higher first cost to the owner.
My DOE2 models usually indicate peak (I mean absolute hourly peak loads)
at about 20% less than Manual J loads, on average, for residential
applications. But even Manual J allows up to 20% above their calculated
loads, which have already been calculated using conservative estimates
for most inputs that are not explicitly defined. The observed
(measured) field results have proven the average residential AC system
to be about 70% to 75% oversized, with some as high as 200% (that's 3
times the peak load). 20% to 25% over ASHRAE's 2.5% design standard is
acceptable to me now, but anything above 25% without some overriding
owner requirement (plans to add on to the current building, etc.) begins
to waste the owner's resources (from first cost to energy and
maintenance costs) and reduce the lifetime of the equipment while at the
same time decreasing his overall level of comfort through limited latent

There! You obviously rubbed a sore spot in my emotional make-up, but I
appreciate the opportunity to make a point. And the point is this: no
matter how good your modeling software is, the outcome is still in the
hands of the user.


Haynes, Glenn's picture
Joined: 2011-09-30
Reputation: 0

That is a good point. It is almost impossible to not oversize the HVAC system from the prospective of designing heating and cooling equiplment. No one can really know what will really happen after the building is built. Actual use in buildings are alwasy different from the design intent as there are big uncertainties for real buildings. Of course it is very important to try to get accurate information during the design phase. But we may more focus on three things,
1: Continue to monior the actual building energy use and equipment efficiency in buildings. These measurements should be very helpful to find some ways to save energy for the newly finished buildings.
2: Improving part load performance in HVAC equipment may be more important than not oversizing HVAC equipment. It seems there is no way to design the very appropriate systems to suit the actual use in buildings. For example, the building loads may be changed a lot from year to year, day to day, hour to hour.
3: For HVAC designer, maybe it is necessary to understand which factors are key variables to size the HVAC system. Maybe for some buildings, the weather conditions (99% or 99.5%) will make big differences in sizing HVAC equipment. But for other buildings, the building use, like internal heat gains, is more important. This can help us which factors we should focus on.


Jin Minming's picture
Joined: 2011-10-02
Reputation: 0

The difference between using equipment sizing software and hourly
building energy simulation software is interpretation of the peak loads.
The equipment sizing software utilizes a worst case scenario of load
contributing conditions to size the equipment and yields an appropriate
peak system load up front - no need to beef it up except to cover errors
or extenuating circumstances.

A building energy simulation code yields the peak loads under the most
probable conjunction of extreme conditions, so you should beef those up
by some reasonable factor if you need to guarantee enough capacity under
the most extreme combination of factors.

Either way, you should end up with approximately the same system size
without unnecessary over sizing. But if you beef up the equipment
sizing software results without proper justification, and then select
the next higher modular equipment size, you will probably end up with an
unnecessarily oversized system; a low personal risk approach, but not
the best design choice for the client who has to pay for it and live
with it.


Haynes, Glenn's picture
Joined: 2011-09-30
Reputation: 0

As a designer of geothermal hvac systems I found this very relevant. For vertical closed loop systems you buy by the foot. Over design of 5 or 10 percent adds that much cost and effects the return on investment for the client. The ground loop is sized based on monthy peak and total loads.

Steve DiBerardine PE CGD LEED AP

Steven DiBerardine's picture
Joined: 2011-10-02
Reputation: 0

Glenn, I couldn't agree more.

I think the advantage of energy modelling as opposed to just pure sizing is
that you can look at the normal operational loads against the peak sizing
loads. This can then allow you to look at modularisation of your plant so
that you only operate as much plant as you need and you run that plant at
optimal efficiency more of the time rather than say running a single boiler
at low part load efficiency. Thus you are not only carrying out your
sizing, but then you are checking whether it will work in practice and
ensure that your design is efficient.

It is important to remember, that a sizing tool which is essentially a
steady state calculation tool will not take the thermal inertia of the
building into account, therefore your building could actually be
hotter/cooler than your cooling sizing calculation suggests it is and this
would obviously effect the sizing of the system. Cooling is clearly very
expensive in terms of capital expenditure, but also lifecycle running and
maintenance costs therefore should be minimised as much as possible.


Paul Carey's picture
Joined: 2011-10-02
Reputation: 0


If this is the procedure you're following for modeling in TRACE, I can see why you view modeling as a pain in the arse. With proper use of schedules, day types, and diversity, what you've described below is easily accomplished in one alternative in one file. I highly suggest attending one of several of our training sessions held throughout the year.

Scott Hintz

Hintz, Scott F's picture
Joined: 2011-10-01
Reputation: 1