[BLDG-SIM] DOE 2.2 experiences?

[James J. Hirsch]

The following paragraphs should help clarify the
issues raised in your message, including:
- Documentation status,
- Ground loop heat pump modeling, and
- The “legal dispute” between J.J. Hirsch &
  Associates (JJH) & LBNL

1. DOE-2/PowerDOE documentation and program status

DOE-2.2 documentation is available, more
comprehensive than for any prior versions of DOE-2,
and complete except for a final review for errors
and changes. The PowerDOE online help file was
completed in early 1997 and includes help on all
DOE-2.2 commands and keywords with the exception of
a small number of mistaken omissions. A draft
version of the DOE-2.2 manuals was produced jointly
by LBNL and JJH and first made available in June
1998 as part of the then current pre-release of DOE-
2.2; LBNL has done a great job in reorganizing and
formatting the scattered DOE-2 documents into a new
set of three manuals. We expect to complete the
final review for errors and changes within the next
three months and make final versions of the
documentation for PowerDOE and DOE-2.2 available at
that time. Draft versions are available now.

Likewise, PowerDOE version 1 and DOE-2.2 pre-
releases are available to anyone and in widespread
use.  Free trial copies, updates, and documentation
can be downloaded upon execution of a no-cost pre-
release testing license. PowerDOE is designated as
pre-release until the final documentation review is
complete. DOE-2.2 is also pre-release for this
reason, and because implementation, beyond internal
test versions, is not complete on some batch
execution features, the metric input/output,
functions, and parametric input commands. This work
is also planned for completion within the next three
months.

Of course, ongoing debugging is an integral part of
supporting PowerDOE and DOE-2.2. Nevertheless, any
existing user would be far better off upgrading to
DOE-2.2 or PowerDOE. DOE-2.2 has better
documentation, fewer bugs, better input and
modeling, better documentation, and runs faster.
PowerDOE has full interactive building description,
simulation execution, and results fetching. Also
available is a developer’s kit which includes: a set
of DOE-2.2 32-bit DLLs, a sample driver program, and
interface documentation to assist in developing
interactive applications based on DOE-2.2
simulations.

2. Ground loop heat pump modeling in DOE-2.1E, DOE-
2.2, and PowerDOE

We first released a version of DOE-2.1E with ground
loop heat pump modeling capabilities in 1995; this
was known as DOE-2.1E version 110.  DOE-2.2 contains
the same basic ground heat-exchanger model with
enhancements to allow the definition of multiple
well/trench “fields.”  When we first released 2.1E
with this feature, we placed a warning into the
manual that we considered this feature in the test
phase; this is no longer the case as extensive
testing and comparisons with measured results has
been completed and show the model to be reliable and
accurate within acceptable bounds.

Davis Energy Group, or their funding agency Pacific
Gas and Electric, could be consulted for more
details on the extensive DOE-2 vs. measurements
comparisons for GLHP systems recently completed. The
most significant issue identified in the testing was
that the model tends to over-predict the daily swing
of the field return temperature during the peak
heating and/or cooling days by as much as 10 percent
(typically, about 5 degrees F.)  Thus the DOE-2
results would tend to cause over sizing of the
field.  The energy predictions, non peak
performance, and longer term (monthly) load
predictions are very good and not significantly
effected by this issue.

Howard McLain, at ORNL, is finishing a smaller,
single building study.  He also identified that the
model does well for overall predictions and also for
the details of temperature tracking during the non-
peak days but also saw larger deviations in field
return temperature predictions for the worst case
peak days.  There are several items about this study
that raise more questions than it provides answers.
Both these studies suggest, however, that we need to
improve the near field (short term) dynamics of the
model before it can be used for sizing if better
than 10% accuracy is required; such accuracy is
normally not an issue but the expense of ground loop
heat exchangers is relatively high so as to make 10%
oversizing undesirable.

Note however, that DOE-2 does not perform default
sizing calculations for well/trench fields; you must
use other means to calculate the number and depth of
bores, etc., and provide this information to the
program. We have been seeking funding to add sizing
calculations but cannot say when they may be
available; so far, it has not been a priority for
funding. One item that is important and different in
2.2 from 2.1E: the specification of circulation loop
min/max alarm temperatures (the temperature range
outside of which attached equipment is not allowed
to operate so as to prevent equipment damage.)  In
2.2, for GLHP systems it is important to specify
these in such a way so as to allow “extended range”
units attached to the loop to operate; for example,
use values of 20 and 130 and they will have no
influence.  This has been the most common problem
that users of DOE-2.2 have encountered; we will need
to implement documentation or default value changes
to help remove this common input error.

The DEG/PG&E work also resulted in our adding
several new features to the DOE-2 GLHP model:
subhour loop operation, cooling towers in parallel
with GLHX field, packaged air-handlers allowed as
GLHP systems rather than just HP unitary systems,
among others.

Included in the DOE-2.2 manuals and the PowerDOE
help is a complete description of the ground loop
heat exchanger model as well as documentation on
WLHP circulation loop and water source heat pump
description options. This information is based upon,
and an expansion of, the DOE-2.1E documentation on
the subject which is divided between the 2.1E
reference manual, the 2.1E Supplement, and the JJH
2.1E-110 enhancements document.  Both PowerDOE help
and the DOE-2.2 manuals unify all the scattered DOE-
2.1E manuals into a single set; a vast improvement.
In the PowerDOE help you simply search for the
"ground loop" phrase, or on a GLHX (or related)
screen just right mouse click on the item and select
item or topic help.  In the DOE-2.2 Command and
Keyword Dictionary you just look in the table of
contents for ground loop heat exchanger, circulation
loop, and system; in the Topics manual just look in
the TOC under system TYPE HP.  If something is
missing from these sections please let us know the
specifics so we can fix the problem.

In summary, there does not exist a fast, easy-to-
use, and inexpensive GLHX model that can do better
than DOE-2 at this time; the DOE-2 model input is
comparatively simple, an annual simulation on a mid-
speed Pentium II is a few seconds, and the program
costs a couple hundred dollars (compared to other
more detailed models that require significantly more
expertise to use, have runtimes from ten minutes to
an hour for a well, and cost thousands to license.)
The more detailed research grade models are very
important tools and should continue to be supported
and developed; their development has lead to
important contributions to the literature that has
been key in the development of the commercially
viable programs such as those based on DOE-2 as well
as those offered by various equipment manufacturers.
Having a range of tools requiring differing levels
of expertise as well as offering varying levels of
detail of analysis is a good situation that will
lead to wider adoption of building energy and
performance modeling within the design community.

A few added comments about user support and the
frustration you expressed.  Software like DOE-2 is
complex due to the large number of options and
capabilities.  We realize that it is often difficult
to fully understand even the best and most robust
documentation. We usually hear from people using
DOE-2 based products when they are having problems;
if they obtained their license from JJH we deal with
them as quickly as possible and if a program bug is
suspected we give the issue priority independent of
their license source. It is impossible to help
people if we are not informed of a problem. DEG was
in constant contact and we dealt with their issues
on an ongoing basis. Many other users contact us
daily with questions and issues; most would say they
get excellent support and issues get resolved more
quickly than with other similar software products.
Of course, we cannot help if we do not know the
issue, the issue is presented as a vague flame, or
competing product advocates make vague statements to
bolster their favored product. We want to help, and
are willing, to a point, to move past your
frustration produced flame but you must provide
specific and detailed explanations of the problems
before we can address them.  The best thing to do is
zip up you input files (.inp and .pdp, and weather
if not ours) and email with a description of the
problem including the exact version of DOE-2 you are
using.  We are, of course, limited by available
resources, but our goal is to provide the best
possible products and to learn from users
experiences and problems to continually improve DOE-
2 and its derivatives.

3. The “Legal Dispute” Between LBNL and JJH

Unfortunately, the information you received from
LBNL was not entirely accurate. As noted above,
documentation is currently available and has been
for some time.

Regarding the “legal dispute,” the real problem is
changed focus at  DOE, and thus LBNL, from DOE-2
development to EnergyPlus development. All
contractual, licensing, and distributions rights to
DOE-2.2 are clearly specified in a contract between
LBNL and JJH and further clarified in a supplemental
written agreement signed by DOE, LBNL, and JJH in
June 1997. These documents provide DOE, LBNL, and
JJH all rights needed to ensure that current and
future versions of DOE-2 will be available, on a
non-preferential basis, to anyone and at a lower
cost than was previously possible.  JJH has and will
continue to strictly follow the letter and intent of
these agreements.

The problem is that these agreements provide the
framework for continuing development, support, and
use of DOE-2.2 (and future versions). Instead,
current DOE program manager is focused on justifying
and increasing annual investment in EnergyPlus. A
constantly improving, expanding, vital, open,
inexpensive and more available DOE-2 (at little or
no expense to the Government) hinders efforts of
this DOE program manager to convince Congress to
continue funding the DOE-2 “replacement” project—at
an ever increasing cost.  For this reason, work LBNL
committed to undertake on DOE-2.2 was not completed.
Further, the DOE program manager would prefer to
hinder JJH’s ongoing privately funded efforts to
develop and deliver new DOE-2 versions—even though
JJH has given DOE and LBNL perpetual royalty free
rights to directly distribute new versions to the
public. Accordingly, a delay in the initial DOE-2.2
release ensued. And, because JJH and its partners
are committed to supporting, improving, and
developing new versions of DOE-2, we are at
loggerheads.

The technology transfer effort for DOE-2 has
otherwise been a model of success, with the private
sector taking the lead to develop and distribute
products like VisualDOE, EZ-DOE, DOE-Plus, PowerDOE,
eQuest, and others. The results of all Government-
sponsored and JJH privately-sponsored DOE-2
development will continue to be available to
developers at minimal cost.  Fortunately, as a
result of this successful transfer, most of the
expertise and all necessary rights for continuing
development of DOE-2 are now equally in the private
sector. Although LBNL’s contributions to DOE-2 have
continued to be excellent, creative, and ground
breaking, their percentage contribution to DOE-2 has
been declining since JJH began its significant
private sector effort in 1988. JJH produced over 60%
of the new code for the initial 2.1E and over 75% of
the new code for DOE-2.2. JJH and its partners and
licensees can and will continue developing new DOE-2
based products, though we regret the loss of what
was a long and productive collaboration with LBNL.

The change in focus at DOE is very disappointing, as
is the DOE program managers attitude that only DOE
can facilitate the development of quality building
simulation software. This attitude is unproductive
and not supported in fact. Although I believe that
the EnergyPlus team is talented and that their
efforts will indeed make some useful and valuable
contributions to building simulation, I also believe
the project would contribute significantly more and
lead to greater advancements in the field if its
HVAC simulation development effort were based upon
the knowledge, expertise, and experience of the
TRNSYS, HVACSIM+, SPARK, DOE-2, and other HVAC
simulation development efforts undertaken during the
past twenty years.

Perhaps posting this response will help clear up the
misinformation. The focus should be on what helps
people make buildings more efficient, and instead we
are mired in DOE and national lab politics.

"John P. Karasaki, P.E." wrote:

> I've noticed a few people mentioning that they are using DOE 2.2.  We have
> used it and PowerDOE on a couple of projects.  Our experience is that either
> can be a bear to use if one is trying to do anything creative since there is
> no documentation for the more advanced features (the reason one would use
> DOE 2.2 vs. 2.1E v.127 in the first place).
>
> We have been particularly frustrated with the ground source heat pump
> algorithms.  After a week of getting inconsistent and unexplainable results,
> a call down to LBNL revealed that the ground loop algorithms are
> "unreliable" and "should not be used."  Instead, they recommended scheduling
> water temperatures, which we eventually did.  The results still seem
> optimistic, but at least the simulation is working somewhat predictably.
>
> I heard that the documentation is done but is held up due to a legal dispute
> between Jeff Hirsch and LBNL.  Unfortunate.
>
> Anyone else have similar experiences with DOE 2.2/PowerDOE?  Any pointers as
> to how to properly define water loops, or where we can get the documentation
> to do so?
>
> Thanks in advance.
>
> BTW, as a 10 year user of DOE2, our frustration with using the undocumented
> DOE 2.2 is akin to what was experienced when I first was learning DOE2.1C!
>
> ======================================================
> You received this e-mail because you are subscribed
> to the BLDG-SIM at GARD.COM mailing list.  To unsubscribe
> from this mailing list send a blank message to
> BLDG-SIM-UNSUBSCRIBE at GARD.COM

--
Jeff Hirsch
James J. Hirsch & Associates
Building Performance Analysis Software & Consulting
12185 Presilla Road
Camarillo, CA 93012-9243 USA
phone: (805) 532-1045
fax: (805) 532-2401
email: Jeff.Hirsch at DOE2.com OR JJHirsch at ez2.net

-------------- next part --------------
A non-text attachment was scrubbed...
Name: JJHirsch.vcf
Type: text/x-vcard
Size: 504 bytes
Desc: Card for James J. Hirsch
URL: <http://lists.onebuilding.org/pipermail/bldg-sim-onebuilding.org/attachments/19990609/a7511bed/attachment-0002.vcf>