|
The default boiler part-load performance curve is
based on the ICES reports (Integrated Community Energy Systems) prepared by Oak
Ridge National Laboratory in the 1970's. This single curve is almost
certainly not representative of all boilers. In fact I would expect that
the part-load performance of atmospheric boilers might be significantly
different than from forced-draft boilers, where the percentage of excess air at
part-load firing can be better controlled.
The default MIN-RATIO is 0.25 for all fuel-fired
boilers. For versions 2.1e and earlier, Fred is correct in stating that
the program holds the efficiency constant at part-load ratios below this
value. Version 2.2/eQUEST/PowerDOE address this issue by applying part of
the START-UP-TIME to the portion of the hour the boiler is cycling.
If you have part-load data that includes the effect
of cycling, then the best approach is to create a curve including the cycling
portion of the range (use the CURVE-FIT command to generate the coefficients),
and set the MIN-RATIO to zero or a very small fraction.
Using a curve-generating routine other than the
2.2's CURVE-FIT algorithm may not give the correct results; DOE-2 expects that
the part-load curve to give a value of exactly 1.0 at a part-load ratio of
1.0, and the CURVE-FIT algorithm ensures this (see the documentation for more
information). If you generate the curve using a different program, such as
Excel, the best fit often does not yield the value of 1.0 at PLR = 1.0
There is a version of the program in development that can use any best-fit
curve; this release will be available later this year.
The fuel-fired boilers do not have default curves correcting
performance for supply and ambient ambient temperature (CAPACITY-FT,
HIR-FT), so the AQUASTAT-T setpoint should have no effect unless you also
define the curves; these hooks were added in case someone had data and
wanted to model this effect.
DOE-2.2 allows you to create libraries of boilers,
chillers, VAV systems, etc.; similar to the material, construction, and window
libraries of 2.1e (store the components in USRLIB.DAT). If you have
performance data for components that you would like to share, why not post them
to this site so that others can include them in their libraries?
----- Original Message -----
Sent: Wednesday, February 12, 2003 4:58
PM
Subject: [BLDG-SIM] Boiler Performance
Curves
Additionally:
First, in DOE2.x, if I'm not
mistaken, the BOILER "MIN-RATIO," or cycling threshold, holds the efficiency
(PLR modified EIR) constant below the specified PLR, contrary to reality and
intuition. I think this input defaults to a PLR of 20%, just a bit above where
boiler efficiency starts to really go down the tubes. So this needs to be
dropped to a very low value(where your EIR-fPLR curve goes wacko) if you roll
your own curves.
Aquastat temp (under BOILER), and HW reset
scheds(HW-LOOP) can also have major effects. Actually you need to create
another custom curve(HIR-FT), and reduce the aquastat minimum, if you want the
HW temp to interact with the boiler, which it most certainly does. See the
DOE2.2 BOILER docs.
Also, if you set up your model with optimistic
assumptions (most defaults), you may miss many hours at very low (<5%)
heating loads.
In any case, most modulating boilers go down to 20 or
25%, and so may not show great savings compared to a 33% low
fire.
A method of calculating part load efficiency for boilers is
outlined in the spreadsheets that can be downloaded from:
http://www.keh-ee.com/cycling.htm
These
require some modification and care in use for purpose of developing DOE2
curves. Steady-state efficiency may vary somewhat with load, producing the
manufacturer's cited increase in efficiency at 50% load. This cited efficiency
improvment may also take into account some assumed HW reset, much as ARI NPLV
chiller points include condenser "relief," at the lower loads.
GOOD
LUCK! In consideration of the other thread: Even with 3D IFC/DXF interfaces,
architects (are any still reading) should not attempt
this.
Fred
Wade McLaughlin wrote:
Hello All,
As part of a comprehensive energy study for a
new elementary school, we are analyzing fuel savings opportunities
for a proposed boiler with fully modulating burners as
compared to a base-case sectional lo-hi-lo boiler. So far, we have
found that, all things being equal, there are no significant savings
associated with such a measure (at least not for the boiler size and models
we've been considering).
We use PowerDOE for most of our energy
simulations. In the course of this study, we have come to question the
default boiler part-load heat-input-ratio curve ( HW-Blr-HIR-fPLR) found in Bdllib.dat. Until now, we
have applied this curve to most base-case boiler calculations (perhaps
mistakenly?). Does anyone know of the origins of this curve or what it
would have been based upon? Most manufacturers we have spoken with
state that peak boiler efficiency occurs between 25% and 50% load.
Using the default curve, peak performance occurs at full
load and efficiency decreases by nearly 10% at half
load.
Also, if this curve is in error, can anyone
suggest a method for developing a more reasonable curve
estimate?
Thanks,
Wade
Wade McLaughlin
DMI
450 Lexington
St.
Newton, MA 02466
p 617-527-1525 x103
f
617-527-6606
======================================================
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
======================================================
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
=====================================================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
|