Você está na página 1de 36

PRV Stability Project Overview

PERF 99-05

Ron Darby, PhD, PE and Abdul Aldeeb, PhD

Presented to the
API 2011 Fall Meeting

November 14, 2011


Los Angeles, CA

API 2011 Fall Meeting - November 14, 2011


Project Objective

To develop a method of predicting the instability or


chatter of a relief valve that is based on sound
fundamental scientific principles, to replace the API
3% Rule.

API 2011 Fall Meeting - November 14, 2011


API 3% Rule

Assumption:
valve will not re-close (or chatter) if the nozzle pressure
does not fall more than 3% of PRV set pressure, due to non-
recoverable friction loss in the inlet line.

Fact:
This criterion is based on static conditions and does not
account for other dynamic factors that affect valve stability.
Thus, it is not sufficient to consider static conditions only for
avoiding chatter.

API 2011 Fall Meeting - November 14, 2011


Program Goals

Develop a mathematical model to predict


stability characteristics (e.g., disc lift vs. time) of
a PRV in vapor/gas service, accounting for
effects that may contribute to instability.
Obtain experimental data on various PRVs for
the purpose of validating the predictive accuracy
of the model.
Incorporate the model into an easily-used
computer program.

API 2011 Fall Meeting - November 14, 2011


Factors Influencing Disc Response

Valve disc dynamic response (damped


spring-mass oscillatory system)
Pressure and fluid momentum forces
on disc
Blow-down setting and backpressure Dd
effects
Dc
Dynamics of fluid response in inlet line
(friction, capacitance, and acoustic
coupling)

Dynamics of fluid response in outlet
d
line
Lift-flow characteristics of the valve
Geometry of flow path leaving the disc

API 2011 Fall Meeting - November 14, 2011


Mathematical Model

The model consists of a number of coupled


nonlinear differential and algebraic equations
which are solved numerically in a spreadsheet.
The output is displayed as the disc lift vs. time,
on a millisecond time scale.

API 2011 Fall Meeting - November 14, 2011


Mathematical Model

Factors considered in the model:

Forces on disc:
Total fluid pressure on disc
Fluid momentum force on disc

Propagating acoustical pressure wave in inlet line

Blowdown and backpressure effects

Dynamic response of disc vs. time:


Spring-mass impulse response
Capacitance of vessel and inlet line

API 2011 Fall Meeting - November 14, 2011


Model Assumptions

Choked gas flow


Valve opens at set pressure (Psg)
Relief mass flow determined at 1.1 Psg=Po
Max lift when Acurtain = Anozzle (choke point shifts from annulus to
nozzle)
Pressure on disc determined by mass flow rate and:
Friction loss in inlet line;
Capacity of vessel and inlet line;
Expansion shock (speed of sound, Li, Di, opening time).
Update force (magnitude and phase) on disc determined by PN, AD,
momentum transfer from fluid to disc (geometry of disc and exit
chamber).
Valve closes when P < PBD
API 2011 Fall Meeting - November 14, 2011
Governing Equations

5 coupled nonlinear equations for disc dynamic response:


x = fn Fx ,t,x,x

Fx = fn PN ,PB ,m
N
PN = fn m N ,t
o ,m
PB = fn m N ,t
o ,m
N = fn x,PN ,PB
m
x= valve lift , Fx = force on disk , t=time

PN ,PB = pressure at nozzle , backpressure


=nozzle mass ra te
= relief mass rate , m
m o N

API 2011 Fall Meeting - November 14, 2011


Model Input Parameters

Valve Design Operating


Parameters Parameters Conditions

1. Mass of moving 1. Inlet piping length 1. Set pressure and


parts and diameter overpressure

2. Spring constant 2. Outlet piping length 2. Vessel pressure and


and diameter temperature
3. Damping of moving
parts 3. Friction loss in 3. Discharge pressure
fittings and temperature
4. Discharge
coefficient (at full 4. Capacity of vessel 4. Relief mass flow
and partial lift) and inlet line rate

5. Geometry of nozzle, 5. Inlet and outlet 5. Gas molecular


disc, and fluid flow piping roughness weight and
path factor isentropic exponent

API 2011 Fall Meeting - November 14, 2011


Stability Criterion

Valve is stable if x(t) is monotonic, or oscillatory with small


decreasing amplitude
0.4
Is this stable lift?
0.3
Stable
Disk Lift

0.2

0.1
0.5

0.0
Time (s) 0.4
0.000 0.005 0.010 0.015

Low frequency

Valve Lift (in.)


0.7
0.3 oscillations
0.6 Flutter
0.5 0.2
Disk Lift

0.4
0.3 0.1
0.2
0.1 0.0
0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
0.000 0.005 Time 0.010 0.015 Time (s)

API 2011 Fall Meeting - November 14, 2011


Stability Criterion

Valve is unstable if x(t) is oscillatory with a large or increasing


amplitude

0.3

0.2
Valve Lift (in)

0.1

-0.1
0 5 10 15
Time (secs)

API 2011 Fall Meeting - November 14, 2011


Factors Impacting Stability

Stability is affected by the following


factors:
damping of the moving parts
stiffness of the spring
opening time of the valve Because the system is
highly nonlinear, all of
blowdown setting these factors interact and
diameter of inlet piping cannot therefore be
considered separately or in
diameter of discharge piping isolation from all the others.
length of inlet piping
Thus it is not possible
length of outlet piping to generalize about the
mass of moving parts effect of any one
"parameter" in isolation
geometry of valve/disc from the others.

API 2011 Fall Meeting - November 14, 2011


Experimental Program

Three different size valves (1E2, 2J3, and 3L4) from


three manufacturers were tested in gas service.
Disc lift vs. time data obtained at 50 and 250 psig set
pressure for each valve.
More than 250 unique test runs were completed with:
No inlet piping - determine opening time
characteristics.
Inlet piping - three different lengths.
Discharge piping two different lengths.

API 2011 Fall Meeting - November 14, 2011


Experimental Program

API 2011 Fall Meeting - November 14, 2011


Experimental Program Collected Data

API 2011 Fall Meeting - November 14, 2011


Testing Observations No Inlet Piping

1E2 @ 50 psig Valves were tested at flow rates


100% Capacity ranging from 10 to 100% of their
measured flowing capacities and all
valves operated without signs of
chattering (high frequency opening
and closing)

3L4 @ 250 psig


80% Capacity

API 2011 Fall Meeting - November 14, 2011


Testing Observations No Inlet Piping

For the same valve, the transient disc


lift response profiles were nearly
identical for most of the tests,
regardless of the pressure rise rate
and the percentage of valve capacity

API 2011 Fall Meeting - November 14, 2011


Testing Observations No Inlet Piping

Mfg. F - 1E2 - 250 psig - No Inlet Piping

0.20

0.16

10% Capacity_1.13 psi/sec


0.12
Valve Lift (in)

24% Capacity_3.20 psi/sec

48% Capacity_6.26 psi/sec

0.08 72% Capacity_9.66 psi/sec

96% Capacity_12.46 psi/sec

96% Capacity_0.96 psi/sec


0.04
72% Capacity_0.72 psi/sec

48% Capacity_0.46 psi/sec

0.00
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
Time (s)

API 2011 Fall Meeting - November 14, 2011


Testing Observations Inlet Piping

18 valves were tested at flow


rates ranging from 30 to 100% 0. 4

of their measured flowing


Run 46 - 30% Capacit y
0.35 Run 47 - 50% Capacit y

capacities.
Run 48 - 100% Capacity
0. 3 Repeat Run 48*

0.25

The range of testing is not

Valve Lift (i n)
significant but results illustrate
0. 2

* The repeat of Run 48 showed that t he


set pressure of the valve had changed

complexity of stability problem.


0.15
t o approximately 232 psig. Testing was
t hen t erminated with this valve.
0. 1

For the same valve at a 0.05

specific inlet piping length, the 0


6.1 6.12 6.14 6.16 6.18 6.2 6. 22 6.24

transient disc lift response Correl ated Time (Secs)

profiles were nearly identical


for most of the tests.

API 2011 Fall Meeting - November 14, 2011


Testing Observations Inlet Piping

44 valves tested valves at 100% capacity:


50 psig valves: 8 of 9 were stable at all tested lengths
with inlet pressure drop ranging from 1.1 to 4.6% of set
pressure (the only unstable case has pressure drop of
4.3%).
250 psig valves: 6 of 9 valves were not tested at 6 foot
inlet lines (presumably because of damage during earlier
testing).
4 valves became unstable at higher inlet piping lengths:
3 of the unstable PRVs had inlet pressure drops < 3% of
set pressure (all were < 5%).

API 2011 Fall Meeting - November 14, 2011


Testing Observations Inlet Piping

44 valves tested valves at 100% capacity:


10 of the 18 tested valves had measured blowdown less
than 7% of set pressure.
5 valves had a difference between blowdown and inlet
pressure drop (w/ 6 foot inlet length) that was < 2% (4 of
the 5 valves became unstable).

API 2011 Fall Meeting - November 14, 2011


Testing Observations Inlet and Outlet Piping

J and L valves set only at 50 psig were tested due to damaged


250 psig valves.
Dynamic backpressure was ~8% of set pressure.
Stability was not impacted by adding outlet piping.

0.25
0.2
No Exit Pipe
0.18 w/ Exit Pipe
No Exit Pipe
0.2 w/ Exit Pipe 0.16

0.14
0.15
0.12
2J3 @ 50 psig
Lift (in)

Lift (in)
0.1

0.1
2-foot Inlet Piping
0.08

0.06

0.05 0.04

0.02

0 0
7 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8 7.6 7.62 7.64 7.66 7.68 7.7 7.72 7.74 7.76 7.78
Correlated Time (secs) Correlated Time (secs)

API 2011 Fall Meeting - November 14, 2011


Testing Program Results

Inlet Piping Length and Pressure Drop


Inlet Pressure Inlet Pressure
Valve Size 2 Feet 4 Feet 6 Feet
Valve Inlet Pressure Drop Drop Measured
& Valve Initial Valve Initial Valve Initial
Number Drop % of Set % of Set Blowdown
Set Pressure Lift Lift Lift
% of Set Pressure Pressure Pressure
(Test) (Test) (Test)
1 1E2 - 50 psig Stable 1.2 Stable 2.0 Stable* 2.7 5.3
2 1E2 - 250 psig Stable 1.0 Unstable 1.7 Unstable 2.2 5.6
3 2J3 - 50 psig Stable 2.2 Stable 3.2 Unstable 4.3 8.0
4 2J3 - 250 psig Stable 2.0 Test Failed 3.0 Not Tested 4.0 7.6
5 3L4 - 50 psig Stable 2.3 Stable 3.3 Stable 4.5 4.3
6 3L4 - 250 psig Stable 1.8 Stable 2.7 Not Tested 3.5 11.9
7 1E2 - 50 psig Stable 1.1 Stable 1.9 Stable 2.6 7.1
8 1E2 - 250 psig Stable 1.0 Unstable 1.7 Unstable 2.3 2.5
9 2J3 - 50 psig Stable 2.3 Stable 3.5 Stable 4.6 9.1
10 2J3 - 250 psig Stable 1.9 Not Tested 2.8 Not Tested 3.8 6.4
11 3L4 - 50 psig Stable 2.3 Stable 3.3 Stable 4.6 5.6
12 3L4 - 250 psig Stable 1.8 Not Tested 2.7 Not Tested 3.6 5.7
13 1E2 - 50 psig Stable 1.1 Stable 1.9 Stable 2.7 9.9
14 1E2 - 250 psig Stable 1.0 Stable 1.7 Stable 2.2 6.4
15 2J3 - 50 psig Stable 2.3 Stable 3.4 Stable 4.5 11.8
16 2J3 - 250 psig Stable 1.9 Not Tested 2.8 Not Tested 3.7 5.6
17 3L4 - 50 psig Stable 2.5 Stable 3.5 Stable 4.6 10.6
18 3L4 - 250 psig Stable 1.8 Unstable 2.7 Not Tested 3.5 4.4

API 2011 Fall Meeting - November 14, 2011


Model Prediction

Valve parameters, design parameters,


and test conditions were utilized as
inputs to the mathematical model for
valve stability modeling.
Dd

All parameters are known or measured Dc


except for fluid deflection angel and
damping factor. These were estimated. d

Generally, the mathematical model was


found to predict the actual test data
reasonably well.

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. E - 3L4 - 50 psig - Large Tank - 3.52 psi/s - 100% of Capacity - 0 ft


Inlet Piping
1.1

1.0

0.8

0.7
Valve lift (in)

0.6

0.5
Model Prediction
0.4 Test Run - Cycle 1
0.2 Test Run - Cycle 2
0.1

0.0
0.0 0.1 0.2 0.3 0.4 0.5
Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. D - 3L4 - 50 psig - Large Tank - 3.53 psi/s - 100% of Capacity - 0 ft


Inlet Piping

0.8

0.7
Model Prediction
0.6
Test Data
Valve Lift (in)

0.5

0.4

0.3

0.2

0.1

0.0
-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5
Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. F - 2J3 - 50 psig - Large Tank - 1.57 psi/s - 100% of Capacity - 6 ft


Inlet Piping

0.6

0.5

0.4
Valve Lift (in.)

0.3
Model Prediction

0.2 Test Data

0.1

0.0
-0.1 0.0 0.1 0.2 0.3 0.4 0.5
Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. D - 2J3 - 50 psig - Large Tank - 1.57 psi/s - 100% of Capacity - 6 ft


Inlet Piping

0.40

0.35 Model Prediction

0.30 Test Data


Valve Lift (in)

0.25

0.20

0.15

0.10

0.05

0.00
0.0 0.5 1.0 1.5 2.0
Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. D - 1E2 - 250 psig - Small Tank - 12.29 psi/s - 74% of Capacity - 6 ft
Inlet Piping

0.15

0.10
Valve Lift (in)

Model Prediction
0.05 Test Data

0.00

-0.05
-0.5 0.0 0.5 1.0 1.5 2.0
Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. D - 3L4 - 250 psig - Large Tank - 11.2 psi/s - 76% of Capacity - 0 ft
Inlet Piping

0.80

0.70 Model Prediction

Test Data
0.60
Valve Lift (in)

0.50

0.40

0.30

0.20

0.10

0.00
0.0 0.1 0.2 0.3 0.4 0.5
Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. D - 2J3 - 250 psig - Large Tank - 6.51 psi/s - 100% of Capacity - 2 ft
Inlet Piping

0.6

0.5

0.4
Valve Lift (in)

0.3

0.2 Model Prediction

Test Data
0.1

0.0
-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5
Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

Mfg. E - 1E2 - 50 psig - Small Tank - 3.02 psi/s - 100% of Capacity Mfg. E - 1E2 - 50 psig - Small Tank - 3.02 psi/s - 100% of Capacity

0.40 0.40
Model Prediction - 0 ft Inlet Piping
0.35 Test Data - 0 ft Inlet Piping 0.35
Model Prediction - 2 ft Inlet Piping
Test Data - 2 ft Inlet Piping
0.30 0.30 Model Prediction - 4 ft Inlet Piping
Test Data - 4 ft Inlet Piping
Model Prediction - 6 ft Inlet Piping
0.25

Valve Lift (in)


0.25
Valve Lift (in)

Test Data - 6 ft Inlet Piping

0.20 0.20

0.15 0.15

0.10 0.10

0.05 0.05

0.00 0.00
0.0 0.1 0.2 0.3 0.4 0.5 0.0 0.1 0.2 0.3 0.4 0.5
Time (s) Time (s)

API 2011 Fall Meeting - November 14, 2011


Model Prediction - Examples

2J3 - 50 psig - Large Tank - 1.57 psi/s - 100% of Capacity - 4 ft Inlet Piping 2J3 - 50 psig - Large Tank - 1.57 psi/s - 100% of Capacity - 4 ft Inlet Piping
& No Outlet Piping & 31.42 ft Oulet Piping

0.6 0.6

0.5 Model Prediction 0.5


Model Prediction
Test Data
0.4 0.4
Valve Lift (in.)

Valve Lift (in.)


Test Data

0.3 0.3

0.2 0.2

0.1 0.1

0.0 0.0
-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
Time (s) Time (s)

API 2011 Fall Meeting - November 14, 2011


Conclusions & Path Forward

PRV stability appears to be affected by several


parameters including valve specifications, system
design, and operating conditions.

Based on the tested valves, mathematical modeling


of valve lift response was possible.

Additional work to test valves of different sizes from


other manufacturers will be required to establish
reliable methods to estimate the unknown
parameters outside the range of sizes and
conditions of those tested.

API 2011 Fall Meeting - November 14, 2011


Questions

Thank you for your attention!

API 2011 Fall Meeting - November 14, 2011

Você também pode gostar