[Equest-users] Oversizing a tower

[Carol Gardner]

Hi Scott,

I'm not convinced that oversizing the fluid cooler would be the best way to
go. It seems like the initial cost would be higher to the client and that
you can achieve what you want with a right sized fluid cooler. If you need a
certain gpm of cooling water to send to the chillers wouldn't the hp to
accomplish that be the same using a larger motor running at part load as a
right sized motor running at full load? I'd suggest using a right sized
fluid cooler with a VFD for part load conditions.

The second strategy is what I know as wet bulb reset. I have attached a full
discussion of it below. It is found in the Volume 2r Dictionary -
Refrigeration Simulation. It is a shame this volume is called
"Refrigeration" because it contains so much more and seems to get overlooked
by modelers who aren't aware of that.

Let me know if you have any other questions.

Carol
Temperature Control in Cooling Towers

<airtofluidflow1.htm>  <minresetplr1.htm>
 Temperature Control in Cooling Towers

The supply temperature that a tower will attempt to achieve is set by the
attached CIRCULATION-LOOP through the COOL-SETPT-CTRL, COOL-SETPT-T, COOL-
RESET-SCH, COOL-SETPT-SCH, and LOOP-SETPT-RNG keywords (see CIRCULATION-LOOP
command). For open towers and fluid coolers only there is one additional
temperature control mode available that tries to achieve an optimal
compromise between tower fan energy and cooling compressor energy. You
activate this sequence by with COOL-SETPT-CTRL = LOAD-RESET in the attached
CIRCULATION-LOOP. The sequence is as follows:

·      The chiller load is determined on an hourly basis. In the field, this
would be accomplished either by a monitoring a signal directly from the
chiller control panel, or, perhaps more commonly, by measuring the
temperature rise across the chiller condenser, and comparing it to the
design temperature rise. For example, if the design condenser temperature
rise is 10F (5.6K), and the actual rise is 6F (3.3K), then the chiller load
is assumed to be 60%. (This assumes that the condenser is constant flow.)

·      The tower air flow (and fan speed) required to reject a given chiller
load is essentially linear with load (open towers) or drops off faster than
the load (fluid coolers). Rather than controlling a variable-speed tower fan
on the basis of leaving tower temperature as is conventionally done, this
algorithm will vary the tower fan speed directly on the basis of chiller
load. For the example given above, the tower fan will run at 60% speed when
the chiller load is 60%. The leaving tower temperature then floats with both
wet-bulb temperature and chiller load. This concept gives priority to
minimizing tower energy consumption, but still achieves chiller energy
savings as the leaving tower temperature floats.

·      To further increase chiller efficiency, it can be recognized that the
tower fan power varies approximately as the cube of the air flow (and load),
and that the majority of the energy savings are achieved when the fan is
close to full speed. Once the fan speed has been reduced to 70% or so,
additional tower fan energy savings will be negligible compared to the
potential chiller savings that might be achieved by reducing the approach to
the wet-bulb.

·      Keywords are provided to allow you to experiment with this concept.
For example, at 50% load the default fan speed will be 50%. You might want
to investigate what happens if the tower fan is not allowed to drop below
65% speed at 40% load.

·      Maximum and minimum condenser temperature limits are included. For
example, you might not want the leaving tower temperature to ever exceed 85F
(29.4C) or to drop below 65F (18.3C). The fan speed algorithm described in
(2) and (3) will be overridden to ensure that these limits are not exceeded.

·      Two-speed tower fans use a variation of this algorithm. As before,
the tower will cycle between low and high speed to maintain the tower
setpoint. However when the leaving temperature drops below the setpoint, the
fan will not cycle between off and low speeds. Instead, the fan will remain
on low speed unless the leaving temperature drops all the way down to
MIN-TWR-WTR-T, at which point it will cycle between off and low to prevent
the tower temperature from dropping any further.

This control sequence is not applicable to single-speed fans. It applies to
both CW loops as well as WLHP loops. Maximum energy savings might be
achieved for WLHP loops, as these systems typically have a lower cooling COP
than chillers, and the heat rejection commonly uses a fluid cooler which has
a higher fan horsepower than an equivalent open tower.

When using the LOAD-RESET control sequence, you can try experimenting with
the values of the MIN-RESET-PLR, MIN-VFD-SPEED, MAX-RESET-SPEED to optimize
the energy efficiency of the system.

The LOAD-RESET tower keywords are illustrated in Figure
36<javascript:TL_7380.HHClick()>and are as follows:
------------------------------

Subtopic help for this Topic:

MIN-RESET-PLR <javascript:RT_1609.HHClick()>

MIN-VFD-SPEED <javascript:RT_1610.HHClick()>

MAX-RESET-SPEED <javascript:RT_1611.HHClick()>

CIRCULATION-LOOP:COOL-SETPT-T <javascript:RT_1612.HHClick()>

CIRCULATION-LOOP:MIN-RESET-T <javascript:RT_1613.HHClick()>
 <javascript:RT_1613.HHClick()> <javascript:RT_1613.HHClick()>
<javascript:RT_1613.HHClick()> <javascript:RT_1613.HHClick()>
<javascript:RT_1613.HHClick()> <javascript:RT_1613.HHClick()>


On Wed, Jul 14, 2010 at 11:37 AM, Scott Hackel <shackel at ecw.org> wrote:

>  Hello –
>
> I’d like to model the energy savings from oversizing a fluid cooler /
> closed circuit cooling tower … My general understanding of this application
> is that the tower can be oversized to create energy savings in two possible
> ways: 1) allowing the tower to run at part load, thus saving fan energy
> assuming a variable speed fan and/or 2) sizing the tower such that the
> average approach during operation is smaller, and therefore condenser water
> temperatures are lower.
>
>
>
> Has anyone had experience with modeling energy savings in eQUEST from an
> oversized tower?  I’ve done this successfully in Trane TRACE, but have not
> had success using the obvious paths in eQUEST – in this case 1) increasing
> the capacity ratio of the tower (from 1 to 1.2 – 1.5) and 2) decreasing the
> design approach (from 7 to 5).  Any thoughts would be helpful, thank you.
>
>
>
> Scott Hackel, P.E., LEED AP
>
>  **
>
> Energy Engineer
>
> *Energy Center of Wisconsin*
>
>
>
> 455 Science Drive, Suite 200
>
> Madison, WI  53711
>
> 608.238.8276  x129
>
> *shackel at ecw.org*
>
> *www.ecw.org *
>
> * *
>
> *The Buildings Team**
> *Design decisions through energy, economics and emissions analysis.
> www.ecw.org/Buildings
>
> _______________________________________________
> Equest-users mailing list
> http://lists.onebuilding.org/listinfo.cgi/equest-users-onebuilding.org
> To unsubscribe from this mailing list send  a blank message to
> EQUEST-USERS-UNSUBSCRIBE at ONEBUILDING.ORG
>
>


-- 
Carol Gardner PE
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/equest-users-onebuilding.org/attachments/20100714/0d5107e5/attachment-0002.htm>