ndamenta s to frontiers

Refrigeration piping for DX systems - Part I

By WILLIAM J. COAD, PE
Because of low first cost advan-
tage-at least a perceived first cost
advantage-there has been a grow-
ing trend to use direct expansion
refrigeration for cooling in variable
air volume systems. This trend has
been accompanied by an extensive
number of refrigeration system
failures, many of which could be
prevented through more careful at-
tention t o t h e basics of refrig-
eration system component match-
ing and piping design.
Most such systems are reason-
ably forgiving regarding air tem-
perature gradients, hunting, etc.,
until these or other problems result
in a condition that destroys the
compressor. Then, of course, the
problem is obvious: T h e com-
pressor has failed! In more cases
than not, the cause of the failure is
not identified, and the compressor
is simply replaced. It is only after
multiple failures that a diagnostic
investigation is launched to deter-
mine if there is indeed a cause ex-
ternal to the compressor itself.
From the standpoint of a building
owner or operator, such an investi-
gation should be launched foilow-
ing the initial failure of any system
in which all elements were not de-
signed and assembled by the manu-
facturer as a standard product.
The three main categories ofcon-
cern in designing or analyzing a re-
frigeration system are: 1) compo-
nent selection and matching, 2)
control scheme, and 3) component
arrangement and piping. This se-
ries of columns will deal with the
category that manufacturers gener-
ally have less control over in field
built-up or split systems and there-
fore should be the first target in
a n y investigation-the refrig-
eration piping system.
In addressing t h i s issue, we
should keep in mind that the two
types of compressor failure are me-
chanicaldamage (usually due to ei-
ther loss of lubricating oil or liquid
slugging) and electrical failure
(usually resulting from extended
overheating caused by inadequate
cooling or extended operation a t el-
evated head pressures). All of these
conditions, in turn, could result
from incorrect component arrange-
ment and piping.
Work input
/
Discharge I ~ n e
Compressor
Heat out
Condenser
From the standuoint of a de-
signer or diagnostic analyst, a ma-
jor concern in system arrangement
should be the range of loading that
can be ulaced on the svstem. The
system k u s t then be designed to
operate successfully throughout
this entire range, If it is determined
that the refrigeration system can-
not operate throughout the entire
range, the design of the VAV sys-
tem being served must be modified
to accombodate the capabilities of
the refrigeration system.
T h e major difference between
VAV svstems and other svstems
such as on-off cooling and-reheat
systems is the requirement for a
broader range of operation. A DX
system in a VAV application is gen-
erally called upon to operate down
to a much lower increment of de-
sign load than one applied to an on-
Suction Ilne
d 2
l
Evaporator Heat In

-
O n this page each m o n t h , t h e author
shares his engineering philosophy b y ex-
ploring a wide variety of topics, ranging
from fundamentals t o new frontiers, as
t h e y relate to building environmental sys-
tems. Mr. Coad is president of McClrrre
Engineering Associates an,d affiliate pro-
fessor of mechanical engineering a t Wash-
ington liniuersity, S t . Louis, Mo. 1 Fiow diagram, simple vapor compression cycle.

100 HeatinglPipinglAir Conditioning February 1990

 off or reheat system. I t is for this
reason that these systems are less
forgiving of poor practice in piping
system design.
Some simple but time-proven
techniques in the design of piping
systems will go a long way toward
preventing compressor failures in
these systems. This series of ar-
ticles will discuss these practices.
A simple DX refrigeration cycle
is represented in Fig. 1, and the ac-
companying simplified pressure-
enthalpy diagram for the cycle is
shown in Fig. 2. The diagram in
Fig. 2 does not account for pressure

n
2-
=
V)
0
LO
2
n

*
Enthalpy, h --Heat ~n--Work+

-Heat out-

2 Pressure-enthalpy diagram, simple vapor

compression cycle.

losses in the piping system. Thus,

each piping segment is shown in
the pressure-enthalpy diagram as a
single state point. Although the
process of pressure drop in the
various piping segments is usually
of little concern, it is of paramount
importance when analyzing the
svstem.
Aside from characteristics of the
piping system t h a t could cause
malfunction or damage to the com-
pressor, improperly designed sys-
tems can also result in excessive en-
ergy losses, so these considerations
will also be discussed in the forth-
coming columns in this series. Next
month's column will address the
design of the discharge and liquid
lines.

Heating/Piping/Air Conditioning February 1990

Fundamentals to frontiers
Refrigeration piping for DX: Condensing pressure bJ/c'
4s
.*
d
Pressure
drop ~n
Part I1 - Discharge line design discharge
l~ne

mance of the condenser establishes

By WILLIAM J. COAD, PE
The discharge or hot gas line car-
ries a mixture of superheated re-
frigerant and oil from the com-
the condensing pressure in a sys-
tem. However, t h e actual head Add~tional
pressure that the compressor sees superheat to
is the condensing pressure plus the be removed
pressure loss in the discharge line.
The effect of the pressure drop in
\energy compressor 1 1 in condenser

pressor discharge to the condenser the discharge line is demonstrated Compressor -4

inlet. The design and arrangement
of this line is normally considered
to be the least critical of any of the
piping sections. However, lack of
proper considerations for some
subtle but significant phenomena
has resulted in serious system
problems. T h e major c o n s i d -
erations in the design of the dis-
charge line are the pressure loss, oil
entrainment velocity, and down-
cycle liquid control.
Pressure loss- T h e perfor-
On this page each month, the author
shares his engineering philosophy by ex-
ploring a wide variety of topics, ranging
from fundamentals to new frontiers, as
they relate to building environmental sys-
tems. Mr. Coad is president of McClure
Engineering Associates and affiliate pro-
fessor of mechanical engineering at Wash-
ington University, S t . Louis, Mo.
in the expanded scale p-h diagram
of Fig. 1.Note on this diagram, that
t o accommodate t h e pressure
needed to overcome the pressure
loss, the compressor must over-
shoot its original target (Point c)
and compress the gas up to Point d.
Since the polytropic compression
curve is leaning to the right, the in-
crease in compressor energy is evi-
dent. Another penalty is that this
results in additional superheat that
m u s t be r e m o v e d i n t h e d e -
superheating stage of the con-
denser, an additional system bur-
den (removing a given amount of
superheat requires much more sur-
f a c e t h a n removing a n e q u a l
amount of latent heat because of
the low gas specific heat of the su-
perheated gas).
Good design practice is to mini-
mize the pressure drop in the dis-
energy with no piping loss
/.c--compressor energy 4
with piping loss
I
1 Pressure-enthalpy diagram demonstra-
ting effects of discharge line pressure drop.
charge line to the lowest practical
value, with t h e maximum limit
being a pressure drop equivalent to
2 F saturation (approximately 3 psi
for R-12 and 5 psi for R-22) for the
entire piping system from the com-
pressor outlet to the condenser in-
let.
Oil entrainment- In all halo-
carbon refrigeration systems using
reciprocating compressors, oil from
the compressor lubricating system
leaves the compressor with the high
pressure superheated refrigerant.
Thus, to prevent total loss of the
conrtnuedon page 119

Site for --,

( a )Single riser (b)Double riser

I (ajLow volume requirement (b)Large volume requirement

2 Oil entrainment techniques for discharge line. 3 Down-cycle liquid control for discharge line.

117
Heating/Piping/Air Conditioning March 1990
Fundamentals to fmntfers
compressor oil, a necessary param-
eter in the design of these systems
is to assure the return of the oil to
the compressor. One technique
that is not commonly used in com-
fort cooling systems today is to em-
ploy an oil separator a t the dis-
charge of the compressor. This
device is designed to separate the
oil from the hot gas and return it
directly to the compressor. The
most common method used in di-
rect expansion comfort cooling sys-
tems is to design the piping system
and all of the vessels to keep the oil
flowing along with the refrigerant
so that none of it is "held up" or
stored in any part of the system ex-
cept the compressor crankcase. In
the discharge line, if the line is run-
ning horizontally or vertically
downward, it is only necessary to
assure that there are no long hori-
zontal dips or low points in the line;
i.e., keep the line and fitting config-
urations always going in the down-
ward direction. If it becomes neces-
sary to rise vertically, two things
must be done. First, the vertical
line must be sized to maintain a
minimum velocity of approxi-
mately 3000 fpm under all capacity
conditions. And second, a t r a p
should be provided a t the base of
the riser to prevent any residual
liquid from draining back on the
discharge valve mechanism on
configurations that are not neces-
sarily common to all systems, it has
received less attention by system
designers. Down-cycle liquid con-
trol becomes a design requirement
of the discharge line when the dis-
charge line rises an appreciable dis-
tance from the compressor to the
condenser a n d t h e riser passes
through a zone or space that could
be a t the system's lowest tempera-
ture on shutdown.
In a refrigeration system, with
the system shut down, the refrig-
erant will condense in the coldest
point in the system. If this happens
t o be the discharge line, the con-
densing liquid will drain by gravity
t o the head of the compressor. Dif-
ferent techniques can be employed
in the design of the system to mini-
mize the volume of the liquid avail-
able (such as pump-down cycles),
but the phenomenon must still be
most common method of pre-
venting compressor damage re-
sulting from this form of down-cy-
cle condensation is to provide a
liquid storage volume in the trap a t
the base of the riser, which is illus-
trated in Figs. 3a and b for a single
riser system. The colored areas
demonstrate the liquid storage vol-
ume available. In the large volume
technique of Fig. 3b, the expanded
volume can sometimes be achieved
by installing the discharge muffler
in this location. If a double riser
such as that in Fig. 2b is used, care
must be taken to put the expanded
volume vessel ahead of the mini-
mum capacity riser connection so
that the oil retention volume is still
limited while the additional refrig-
erant liquid volume is available.
In next month's column, we will
discuss the design of liquid piping
systems.

In systems that have cylinder

unloading for capacity control or
multiple compressors that can cy-
cle on a common piping system,
multiple risers may be required as
shown in Fig. 2(b), where t h e
smaller riser is sized a t the required
velocity for the minimum flow ca-
pacity of the system and the other
riser is sized a t the design capacity
less the minimum capacity. The
colored zone in the larger trap illus-
trates the maximum volume of oil
that the trap will retain following
extended operation a t minimum

Down-cycle liquid control-

Since this phenomenon is not in-
herent in refrigeration piping sys-

Heating/Piping/Air Conditioning March 1990

Fundamentals to frontiers
Refrigeration piping for DX:
Part IV - suction line design
By WILLIAM J. COAD, PE
In last month's column, we
stated that "the liquid line is per-
haps the least critical of the piping
segments insofar as potential dam-
age t o the compressor is con-
cerned." The antithesis of t h a t
statement is that the suction line is
perhaps the most critical. The suc-
tion line is the last leg of the jour-
ney of the refrigerant and oil after
leaving the discharge valve to re-
turn to the compressor.
The two objectives in the design
of the suction line are to hold the
pressure loss to the minimum pos-
sible value and to return the oil
(which separated from the liquid
refrigerant as it changed to the gas-
On this page each month, the author
shares his engineering philosophy by ex.
ploring a wide variety of topics, ranging
from fundamentals t o new frontiers, as
to sys-
tems. Mr. Coad is president of M c C l u ~ e
Engineering Associates and affiliate pro-
fessor of nzechanical engineering a t Wash-
ington Uniuersity, S t . Louis, Mo.
HeatinglPipinglAir Conditioning May 1990
Considering first the pressure
drop, refer to Fig. 1, a partial scale
pressure-enthalpy diagram that il-
lustrates the effect of suction line
pressure drop on the compressor
energy. As the gas leaves the evapo-
rator with about 10 F of superheat
at Point a, isentropic compression
would follow the isentropic Curve s
to Point b. However, the actual
compressor curve would lean out to
the right along the polytropic line
represented by the curved Line a-c.
However, a pressure drop in the
suction line would be represented
by the vertical downward Line a-a'
so that the refrigerant enters the
compressor at the lower Point a'.
Since compression lines flatten out
moving upward, this is seen to have
a somewhat magnifying impact
derstand the impact of the pressure
drop upon the power or energy re-
quirement in each particular sys-
tern, but as a guideline, common
practice has been to limit the pres-
sure loss in the entire suction sys-
tem to a pressure differential
equivalent to 2 F saturation tem-
R-22,z psi(approximately
perature for R-12). 4 psi for
The second objective, the return
of oil to the compressor, requires
practices similar to those discussed
for the discharge line. In the suc-
tion line, however, the problem is
more critical since the oil is more
viscous and therefore harder to
move a t the lower temperatures.
Starting a t the outlet of the
evaporator, the suction line should
always drop vertically downward
or run horizontally at a downward
slope to a point where it then drops
vertically downward. The objective
is to clear the evaporator of any oil,
t h u s preventing i t from being
trapped in the evaporator. Once
the oil has been cleared from the
evaporator, the only remaining
problem is to keep it moving. The
velocity in horizontal lines is gener-
ally adequate to move the oil along
with the refrigerant effectively.
However, if the situation allows, it
4/s in. per ft) to assure that there will
be no accidental low spots in the
lines. Also, if there are any branch
connection inlets into the horizon-
tal line, such as a branch inlet from
another evaporator, they should al-
ways enter the top of the line.
page 9.5
conti~zuerIo~i
93
Fundamentals to frontiers
( O I I ~ I I I I I L ~ ~ \ ~9
O I I ~ ~ ~ I ~ L
Where the line must rise in the
direction of flow, a trapping tech-
nique similar to that for the dis-
charge line is used, as shown in Fig.
2. In a system with a single com-
pressor and no cylinder unloading
(or other type of variable flow), a
single trap is used, as illustrated in
Fig. 2a. For the suction line, the
vertical line should be sized for a
minimum of 1500 fpm and a max-
imum of 2500 fpm.* As with the
discharge line, if the system is de-
signed for variable refrigerant flow,
such that a t the lowest load or flow
condition the velocity in the riser
* T h e ~elocrt\ ranges g w e n ~ r zthcc serles
h a w been demonstrated t o prouzde effec-
tcve ocl return wrthout excecclue pressure
drops Howeuer, m u c h more e x t e n w e data
are auazlable relatcng t o this matter In t h e disaster. ~
A S H R A E Handbook chapter " S ~ s t e m
Practtce, for Halocarbon R e f r ~ g r r a n t s "
(currentlv Chapter 3 of t h e 1986 Refrig- ple and follow the basic objectives
eration Handbook
Heating/Piping/Air Conditioning May 1990
would drop below the minimum re-
quirement, then the double riser
technique of Fig. 2b is used, with
the minimum capacity riser sized
for nominally 1500 fpm a t the mini-
mum flow capacity and the other
riser sized for the maximum less
the minimum capacity.
Needless to say, the successful
design of a refrigeration system re-
quires extensive knowledge of all of
the aspects of the systems, such as
component capacity matching,
controls, and specialty devices, in
addition to the piping systems
themselves. But an understanding
of proper piping practices is essen-
tial if the designs are to be success-
ful, and the addition of controls or
devices to offset poor pip-
ing practices, although all too com-
mon, is a technique that will lead to
hthe designer
~ ~ should
,
keep the piping 'ystern sim-
discussed in this series. Q