ASHRAE Journal - November 2013 - 92

2013 ASHRAE TEcHnology AwARd cASE STudiES

to be modeled within TRNSYS. To achieve this, the
installation validating the model as accurate. The
design team used a DOE2-engine building model called
2005 design was based on ASHRAE/IESNA Standard
TRNBUILD.
90.1 - 2004.
The TRNBUILD application includes a radiant floor
Indoor Air Quality
component, which calculates the average surface
Because proper humidity control was critical to the
temperature of the radiant slabs. This was extremely
Walmart's store design, the designer, in conjunction
important to the modeling of the slab losses as the surface temperature drives the heat transfer rate to the soil with the manufacturer, developed dehumidification
air handlers using heat pump technology specifically
below the slab.
for this project, permitting the control of the store temAs the heat transfer characteristics of massive
perature, humidity, CO2 demand control ventilation,
slabs are not modeled within typical DOE2 engines,
and building pressurization. Heat pump technology
including TRNBUILD, it was necessary to model the
heat loss from the building to the ground. A custom- with heat rejection from the compressor was used for
the reactivation cycle, unlike with traditional dehuized ground model was integrated to the slab surmidification systems where gas is
face to estimate the losses/gains
used for the reactivation cycle.
to the slab. However, this model
Table 1 Monthly energy use.
Because dehumidification
necessitated many hundreds of
ElEctricity
UsE (kwh)
2008-2009
Billing Days
controlled four major control
soil nodes to simulate the heat
parameters of the building, the
transfer through the slab and into
DEc./Jan.
39
333,519
sequence had to be prioritized.
the soil beyond. This requirenov./DEc.
28
250,296
Humidification became the first
ment increased the model time
oct./nov.
25
209,623
priority where it was critical for the
substantially.
sEpt./oct.
32
280,332
floor slab not to create moisture at
The energy model required more
its surface and to ensure that the
than 24 hours of computing time on
aUg./sEpt.
30
289,198
case glass of refrigeration units did
the fastest machine available to perJUl./aUg.
33
321,005
not condensate. It was important to
form one year of model time. This
JUn./JUl.
30
269,694
the client that customers can view
compares to a typical model that
the products inside the case.
requires, at most, one hour of comMay/JUn.
28
239,033
Temperature was the second priputing time per model year.
apr./May
32
259,682
ority where it worked in conjuncAs heat from the radiant slabs
Mar./apr.
31
238,200
tion to the slab radiant flooring
slowly increased the soil temperaFEB./Mar.
29
240,284
operation. The third priority was
ture beneath the building, it was
CO2 control for ventilation, and
necessary to perform multi-year
Jan./FEB.
31
277,829
last was pressurization to minimize
runs to analyze the change in slab
DEc./Jan.
35
227,144
infiltration through the main slidlosses after long-term operation of
nov./DEc.
22
202,114
ing doors.
the building. In normal modeling
Air quality during construction
applications, the model is tuned
and pre-occupancy was also taken into consideration.
and validated by running the model for the necessary
Although the project was not LEED certified, certain
time to measure each effect. In this case, five years
LEED guidelines were followed where air quality was a
proved sufficient to measure the long-term perforfactor and where the constructor followed guidelines as
mance. Due to the long computing time, the iterative
part of their standard practice. To reduce the source of
model verification process took months of real comcontaminants, the building was continuously purged for
puting time.
one month prior to occupancy, while the store racks was
The energy model predicted a reduction in energy of
being installed.
57%, compared to the actual reduction of 60%. FortyThe project team also provided commissioning agent
four percent of the predicted reduction was electrical,
services, providing site investigation and site support to
in comparison to a base model 2005 design, which
the mechanical and controls contractor on a daily basis.
resulted three percentage points lower than the actual
92

ASHRAE JouRnAl

ashrae.org

N ovem ber 2013

ASHRAE Journal - November 2013

Table of Contents for the Digital Edition of ASHRAE Journal - November 2013