Hydrocarbon Processing - May 2021 - 47

Maintenance and Reliability
from the starting temperature of 20°F for
a minimum outlet temperature of -70°F
at the high-velocity point. This requires
ANSI B31.3 pipe stress test analysis for
typical carbon steel (e.g., A106) or a different metallurgy selection (e.g., stainless
steel). It was found that the A106 piping
at this system's pressure was adequate,
but the manual block valves' internal soft
goods were only adequate for -50°F. A
change in design was required for this system's valves. In addition, thermal shock
for the sudden change in temperature was
evaluated and found not to be an issue but
should also be considered in designs.
Case 3: Operating valve. A different city gas supply was utilized at 450
psig at inlet temperatures of 20°F-40°F
in the winter and 100°F in the summer.
The system regulates pressure 40 psig-60
psig downstream through a 1-in. regulator. Operations measured 1 MMsft3d gas
flow. Ice forms on the system's 1-in. outlet connection almost year-round, baffling operations and engineers at the site
in the summertime where the simulation
showed that the valve outlet temperature
should be typically 40°F or higher. Isen-

Case 4: PSVs/flare headers. Most
PSVs generally have a Mach velocity
across the throat of the PSV, and Mach
velocity is common in tailpipes. Some
engineers avoid Mach on headers for
acoustic vibration reasons. Others try to
avoid Mach by wrongly assuming it is a
design or code compliance constraint.
In addition, the temperature at the tees
is seldom examined. In multiple cases, it
40

T° = 120

T° = 80

T° = 40

T° = 0

5
0

0.2

0.4

0.6

0.8

Isentropic ΔT, °C

Case 2: Retail city gas to flare. For

one client, retail city gas is injected at 440
psig into a flare at below 5 psig to meet new
regulatory Flare RSR (40CFR63.670) requirements. The gas can reach 20°F at 440
psig in the winter, and the valve adiabatic
temperature loss is approximately 40°F.
The outlet velocity can reach about 0.9
Mach. The system utilizes normal carbon
steel piping and the outlet temperature
is roughly -20°F by accounting for the
JT effect and as per simulation results.
At a velocity of 0.9 Mach, the isentropic
change in temperature was found to be
50°F, resulting in a 90°F temperature drop

tropic effects and comparisons to simulation data are presented in TABLE 2 at the
valve outlet connection.
Due to low-temperature results with
lower pressure or lower inlet valve temperature, a mechanical stress test study
was required. The existing piping system
downstream of the regulator was not designed for these conditions and the system was routinely below the MDMT of
the metal, posing a risk to failure longterm. The regulator size was increased
to 2-in. to eliminate any 1-in. section of
piping and decrease the velocity, thereby
increasing valve outlet temperature. The
resulting operation showed no more icing
in the summer and operating temperatures above the MDMT in the winter.

Isentropic ΔT, °F)

T ≈ 63.8°F.
The calculated temperature through Eq. 2
is slightly lower than the value found by the
FLIR camera but is within reason due to
uncertainties in the flow measured (orifice
meter with known drift) and heat transfer
through the wall of the pipe. The principles
applicable to Eq. 2 accurately represent the
steady-state system of interest.
When to consider isentropic effects?
Isentropic effects are relevant for numerous systems in industry, including control
valves, pressure safety valves (PSVs), flare
headers, compressor station designs and
others. It can also potentially explain failures in a system due to exceeding the minimal design metal temperature (MDMT)
when conventional modeling does not
predict colder conditions. On a practical
level, it can also explain icing on the outside of piping when modeling predicts
higher than freezing temperatures. The
relationship of isentropic change in temperature with increased Mach velocity for
different k-values can be found in FIG. 3.
The k or Cp /Cv values for a list of fluids
can be found in TABLE 1. When a fluid is
examined with various starting or stagnation temperatures, the resulting isentropic change in temperature varies. In FIG. 4,
a fluid with a k-value of 1.3 is examined
over a range of starting temperatures.
Why consider isentropic effects?
Consideration of the design of valves
(both the metal body and the internal
soft goods), thermowells and pipe wall
material are important considerations in
industry. For example, soft goods within valves (e.g., PSVs and manual block
valves downstream) may not be able to
endure the low temperatures that occur
from isentropic effects, as demonstrated
in the next two cases.

Mach velocity
FIG. 4. Isentropic ΔT (°F or °C) vs. Mach velocity at k-value of 1.3 for various starting temperatures.

TABLE 2. Outlet temperature vs. isentropic effect temperature of the system with
different operating pressures and starting temperatures
Inlet
Simulation predicted
temperature,
outlet adiabatic
Pressure,
°F
temperature, °F
psig

Mach
velocity

Actual valve
Isentropic, temperature,
∆T, °F
°F

Case 1

0.338

Case 2

-22

0.388

-31

Case 3

-25

0.44

-39

Hydrocarbon Processing | MAY 2021 47

Hydrocarbon Processing - May 2021

Table of Contents for the Digital Edition of Hydrocarbon Processing - May 2021