[Bldg-sim] HW loop losses

[Fred Porter <fporter@xxxxxxxxxxxxxx>]

All;
In an earlier posting Steven Gates wisely urged us to account for steam and
HW loop losses and wrote:

Hot-water loops can also have very large losses; I have found that the losses in VAV/hot-water reheat systems typically exceed the annual reheat load.

I'm curious if this is by measurement or simulation? Are "the losses" just
the piping heat losses, or does that include boiler losses and valve leakage?
I presume this observation applies (as a "typical" finding) mostly to California
and other moderate climates. Does anyone else see evidence of this as a "typical"
situation?
  
Recently I've had the opportunity to compare actual heating, and reheat loads
with simulated loads for both HW and resistance heating systems. Perhaps
I got a little sloppy in some of my simulation assumptions, but in every
  case, measured heating loads were greater than simulated (including
resistance reheat systems without loop losses). In one case, of incredibly
poor new building envelope detailing and optimistic simulated building characteristics,
the measured heating use was 5-10x the simulated values. Because I've seen
this under-prediction in data from electric reheat, I've mostly assumed the
simulation problems were in the zone characteristics, but the quote above
makes me wonder if I also need to pay more attention to the loop and boiler.
In most of these comparisons, the real cooling energy was close to the simulated
cooling energy or slightly over-predicted. 
  
Many assumptions about particular components, arrangements, and control schemes
can cause under-prediction of both reheat needs and loop losses, while having
smaller or negative effects on cooling energy.
Pardon the somewhat non-parallel structure. For the most part, the actual/real
situation is described first.
  
Including:
Envelope 
Walls that "the wind blows through," in reality, but not in the model
Even in better buildings, the ASHRAE 90.1 model assumption of zero occupied-period
infiltration is not a valid assumption for real building pressures, and characteristics
(lobby traffic, non-curtainwall construction, wind over 10 mph, .... ). This
is just for comparison purposes.
Walls with lots of funky exposed steel penetrations, headers, columns that
somehow didn't get averaged into the already miserable modeled steel stud
wall R-values.
Real above grade slab edges without any functional insulation that were modeled
as underground, or insulated, or both, or ignored.
Real blinds are often deployed for glare control, reducing solar gains that
offset modeled heating loads, and may not be modeled in "90.1" models
Models often leave out shading from other buildings, terrain, and landscaping.
In DOE2, the simple shading coefficient method over-predicts actual solar 
heat gains in high-incidence angle (winter E&W) situations. 
 Because of the proportionally higher losses from the greater surfaces, corner
zones may have a substantially higher balance point temp (so the heat is
on below a higher ambient) than "thermal blocks" modeled with one exposure's 
glazing only. This may significantly increase raw heating loads, and will 
increase loop losses from a lengthened boiler and HW loop operating season.
  
Internal Loads 
Actual W/sf plug loads are not the same over the building, though they may
get modeled that way for simplicity. Many of the larger plug loads (printers,
copiers, servers, pantries...) tend to be clustered in core areas.
 Actual building occupancy and associated heat gain is usually much lower
than the design occupancy. 
 Even in non-"daylighting" real buildings, lights and their heat gains, may
be off in a significant # of perimeter zones, though a model has a schedule
of 90% on all day.
  
Zone control
Winter supply air temps cooler than assumed.
Overall supply air temps cooler than assumed to satisfy one zone that is
starved for air, or overloaded with electronics.
Minimum perimeter primary airflows higher than assumed or specified.
Thermostats installed on exterior walls.
  
Plant & Loop losses
No reset of HW loop temps, even during very mild weather, contrary to assumption.
  
Giant atmospheric boilers with constant primary HW flow, operating at 2% load
to serve one or two zones for reheat. (DOE2 users need to be careful, because
modeled boiler efficiency is held constant below the cycling limit, in reality
it can get much worse.)
HW loops operating at such high differential pressures that leakage through
heating coils is almost assured. This can't even be modeled in DOE2.
Real piping losses obviously greater than those modeled with adherence to
the ASHRAE 90.1 ECB method instruction that "Piping losses shall not be modeled
in either building model." In the ASHRAE methodology, if the proposed building
has HW heat, then the budget building also does. However, these assumptions
can get carried over into overall building energy cost projections and comparisons,
when one is in a hurry, or does not want to explain to the clients why their
real building is projected to use more energy than the "minimally-compliant
code" building.
  
These are often characteristics that cannot be directly transferred from
plans to simulation inputs. Some are things that can't be known in advance,
at least not with precision. Many can and should be fixed after the fact,
or avoided with proper detailing and specification. Some just require a little
careful modeling thought.  
-- 
Fred W. Porter
Senior Engineer
Architectural Energy Corp.
2540 Frontier Ave. Suite 201
Boulder CO 80301
email: fporter@xxxxxxxxxxxxxx 

  
  


======================================================
You received this e-mail because you are subscribed 
to the BLDG-SIM@xxxxxxxx mailing list.  To unsubscribe 
from this mailing list send a blank message to 
BLDG-SIM-UNSUBSCRIBE@xxxxxxxx