Automation & Control

Instrumented overpressure protection cuts costs, eliminates discharges

An instrumented overpressure protection system (IOPS) offers significant cost savings and does not discharge to the environment. Effective applications depend on dynamic fluid flow characteristics and shutdown valve response time in isolating the pressurizing source. Transient analysis can verify IOPS effectiveness.

Tek Sutikno, Fluor Corporation, Houston

The use of an instrumented system for overpressure protection differs from the use of conventional, spring-loaded relief valves. These valves typically discharge to the atmosphere or to a flare system. For relief valves in hydrocarbon service, a flare system is typically required to combust the hydrocarbons from relief valve discharges. In addition to the capital cost associated with the flare system, flaring can lead to the atmospheric emission of acid gases and may involve applications for special permits from the environmental authority. When the system protected by a relief valve is in sour hydrocarbon service, flaring sour gas is strictly regulated in general, or prohibited in areas where facilities to reduce or eliminate sulfur oxide (SOx) emissions are mandatory.

An instrumented overpressure protection system (IOPS) typically involves an automatic shutdown valve that is capable of quick, reliable isolation of the high-pressure source to protect equipment from overpressurization. Isolation of the high-pressure source eliminates the need for any vent discharge and associated flare system. To enhance reliability, IOPS is typically designed with certain levels of redundancy, sometimes referred to as a high-integrity pressure protection system (HIPPS).¹

Oil and gas industry usage of IOPS has been increasing.² This is due primarily to IOPS cost-effectiveness and the absence of IOPS environmental impact, relative to a conventional pressure relief system. For some projects, advantages from IOPS applications can be the decisive factors for project viability. However, an IOPS system can be installed only in certain overpressure protection applications.

While US codes may not permit substitution of a spring-loaded relief valve with an IOPS in certain applications (such as pressure vessels), the US Department of Transportation (for example) allows IOPS usage for overpressure protection of pipelines in defined categories.³ In general, IOPS applications are limited to pressure letdown stations, where the high-pressure source needs to be isolated quickly to protect the low-pressure-rated discharge section from overpressure.

An IOPS generally relies on an automatic shutdown valve/ device to quickly isolate the high-pressure source and prevent the protected pipeline segment or equipment from overpressurizing. Because automatic shutdown valves require finite times to close, it is extremely important to confirm that the shutdown valve closing time is adequately small to assure IOPS integrity for preventing the pressure of the protected system from exceeding design pressure, plus the allowed accumulation.

This confirmation requires a transient evaluation of the IOPS and the associated system it protects. This article presents an evaluation method that designers may consider as an option for evaluating dynamic or transient characteristics of an IOPS installed on an upstream production flowline from a well producing gas-saturated crude oil. This method also can be considered for analyzing an IOPS, for any pressure letdown station in a downstream processing unit where the applicable codes and regulations allow proper use of an IOPS for overpressure protection. Constraints for IOPS applications are discussed in the literature.^1,7,8

IOPS

Fig. 1 shows an IOPS application that includes an automatic shutdown valve (SDV) and a redundant system of pressure instrumentation for closing the SDV, if high pressures occur in the protected segment. The protected segment downstream of the SDV is designed for a lower pressure rating than the upstream segment, including the wellhead. When the pressure downstream of the SDV reaches the set point, this valve will close.

Fig. 1. IOPS system sketch.

A high-pressure event will occur primarily in one of two scenarios. The first one is when flow is blocked, due to either the closing of a downstream block valve (BV) or other causes, such as a hydrate formation. The second scenario is when the upstream regulating choke valve (CHV) fails to control the pressure. In this example application, the CHV regulates the well’s fluid production rate by adjusting the valve opening. Pressure downstream of this choke valve is mainly determined by the pressure drop in the production line and the operating pressure of the downstream receiving unit.

TRANSIENT ANALYSIS

The transient pressure surge profile of an IOPS varies with the fluid flow characteristics of the high-pressure source, in addition to the BV and SDV closing times. In the IOPS example shown in Fig. 1, the high-pressure source that forces flow to the protected segment is the well reservoir pressure, which is at 9,000 psig in the well’s early production period.

Pressure downstream of the choke valve operates at 1,700 psig, and the SDV begins to close when this pressure reaches 1,810 psig, which is 6.5% above normal operating pressure. Because of the high-pressure drop from 9,000 psig to 1,700 psig, and the liquid flashing in the low-pressure region, the choked flowrate through the CHV remains unchanged at a constant inlet pressure. This is true, as long as the downstream pressure does not increase to roughly half of the inlet pressure, or 4,500 psig.

In the event of downstream flow blockage, fluid pressure in the protected segment begins to increase, primarily because of two factors. The first factor is the accumulation of mass flow from the choke valve to the blocked segment. The second factor is the conversion of kinetic energy to pressure in the segment, when the outlet flow is blocked, and is generally referred to as “water hammer.” The accumulation of mass in the protected segment can be expressed as:

Eq. 1

V is the total available volume of the protected segment between SDV and BV in cubic feet; r is the overall fluid density in the segment in lb m/cu ft; t is time in seconds; and m is the net accumulation rate in lb/sec of mass in the segment. If the closing time of BV is CT_BV, m varies within this CT_BV time period and can be written as:

Eq. 2

The primary goal of this analysis is to evaluate whether the protected segment, at a given SDV closing time (CT_SDV), will reach a pressure higher than the design rate, or MAWP, plus the allowed accumulation. The more that CT_BV increases, the longer that CT_SDV can be allowed to go on without exceeding the protected system’s overpressure limits.

For cases where CT_BV is shorter than the time required for the protected segment rate to increase from normal operating pressure (1,700 psig) to the high-pressure set point (1,810 psig) for closing the SDV, BV is completely closed when SDV begins to close. Then, m_out = 0 in Eq. 1, and this equation can be integrated from time, t = 0, to t = CT_SDV. The initial density, r 0, corresponds to the fluid density in the protected segment when the pressure reaches high set point.

The choked flowrate (m) through the CHV remains constant, as long as inlet pressure does not change, and the downstream pressure does not increase to a level where the choked (critical) flow changes to a sub-critical flow. However, because the SDV is closing during the time period CT_SDV, the CHV downstream pressure (or the SDV inlet pressure) will begin to increase at some point during the SDV closing time, and the flowrate (m) will decrease and stop at t = CT_SDV.

As an example, when the CHV is open at a valve flow coefficient (C_V) of 5.5 (corresponding to the production rate setting) and the SDV (which has a maximum C_V of 248) closes with a CT_SDV of 5 seconds, a plot of m versus the closing time (CT_SDV) can be developed, as shown in Fig. 2. This plot can be numerically integrated to calculate the final density (r _f) when SDV completely closes.

Fig. 2. Mass flow vs. SDV closing time.

The calculated rf can be correlated on a constant enthalpy line to find the protected segment pressure at t = CT_SDV. The flow through SDV is essentially isenthalpic, because the flow does not involve work, and elevation change and heat loss are negligible in this case.

Fig. 3 shows the protected segment’s pressure surge profile during SDV closing time (CT_SDV). At an 1,810-psig design pressure for the protected segment and a 10% allowable accumulation, the SDV closing time of 5 sec does not allow the pressure inside the segment to exceed the 1,991-psig limit. However, the second source of pressure from the conversion of kinetic energy has not been included in Fig. 3 and will be discussed later.

Fig. 3. Pressure surge profile, CV = 5.45.

Fig. 4 shows another case where the CHV is open at a C_V of 10 and the SDV C_V remains the same at 248, except that its closing time increases from 5 seconds to 10 seconds. As shown, the SDV closing time of 10 seconds exceeds the overpressure limit of 1,991 psig. This also illustrates that the opening maximum C_V’s of the flow regulating a CHV or other control valves upstream of the SDV need to be considered when evaluating IOPS effectiveness.

Fig. 4. Pressure surge profile, CV = 10.

As discussed earlier, the second source of overpressure is the conversion of kinetic energy, or velocity head, to pressure and acoustical energy. Time-dependent profiles of this pressure surge, due to kinetic energy, will involve a complex mathematical analysis for which the initial and boundary conditions are difficult to verify. However, maximum pressure rise from the transformation of kinetic energy in a blocked discharge case of a pipe segment has been investigated^5,6 and can be shown as Eq. 3:

Eq. 3

In Eq. 3, Du is the fluid velocity change before the impact of valve closure, and c is the sonic wave speed that can be expressed as Eq. 4:

Eq. 4

In Eq. 4, K is the bulk modulus of fluid elasticity, d is pipe diameter, e is pipe wall thickness and E is the Young’s modulus of elasticity for the wall material.

The value of c calculated from Eq. 4 corresponds to the wave speed for pipe anchored with expansion joints throughout, and it varies depending on the support arrangement.^4,6 Eq. 3 calculates the peak pressure surge when the block valve (BV) closing time is less than the critical time or wave cycle time defined as 2 L/c, where L is the length of the pipe between BV and SDV.

The pressure surge due to transformation of kinetic energy will be less when the BV’s closing time is longer than the wave cycle time. Additionally, the presence of gas or vapor in the pipe significantly reduces the wave speed (c) and the magnitude of the peak pressure surge.⁴

For the worst case where CT_BV is shorter than 2 L/c, Fig. 5 shows the potential pressure surge of the protected segment when the water hammer effect is included. As shown, an additional peak pressure increase of about 176 psi can occur during CT_SDV after the BV is closed. The allowed overpressure limit (1,992 psig) is exceeded in less than one second for the case where CT_SDV is 5 sec and the opening C_V of CHV is 5.5. This illustrates the need to revise pertinent IOPS design parameters to account for this pressure surge. Dependent upon the specific application, potential revisions of IOPS parameters include lowering the high-pressure shutdown set point, increasing the BV closing time, decreasing the C_V of the flow-regulating valve (if viable) and increasing the protected system’s surge volume.

Fig. 5. Pressure surge profile with hammer

Pressures shown in Fig. 5 are the momentary peak pressures along the protected segment. These transient pressure surges from kinetic energy transformation result in vibration and noise generation, sometimes extremely violent. Fluid friction and imperfect elasticity of both the fluid and the pipe wall help to dampen out the vibrations. Fluid pressure in the protected segment eventually reduces to the level shown in Fig. 4, which indicates the average or final settled/ static pressure of the protected segment.

As shown in the above example cases, transient pressure surges need to be analyzed to assure IOPS effectiveness, in addition to conducting a separate study for evaluating IOPS reliability as a function of instrument redundancy, component reliability and testing frequency. Flow characteristics and fluid thermodynamic properties, including phase behavior, are typically required to perform the dynamic analysis.

For a single-phase system within the sub-critical flow region, the analysis can become less complex. In most, if not all IOPS cases, the required analysis is a worthwhile effort when the benefits of significant cost savings and the absence of environmental impacts are considered. 

LITERATURE CITED

  ¹  Summers, A. E., “Consider an instrumented system for overpressure protection,” Chemical Engineering Progress, November 2002.

  ²  Chun, E., “PG&E improves system reliability,” Pipeline and Gas Technology, April 2003.

  ³  Code of Federal Regulations, (CFR), Title 49, Part 192, Office of Pipeline Safety, Research and Special Programs Administration, Department of Transportation, Oct. 1, 2002.

  ⁴  Wylie, E., and V. L. Streeter, Fluid Transients, McGraw Hill, 1978.

  ⁵  Nayyar, M. L., Piping Handbook, McGraw Hill, 2000.

  ⁶  Zaruba, J., Water Hammer in Pipe-Line System, Elsevier, 1993.

  ⁷  Gruhn, “Safety instrumented system design: Valuable lessons learned,” Hydrocarbon Processing, August 2000.

  ⁸  Van Beurden, I. and R. Amkreutz, “Safety integrity level verification – A PFD average calculation is not enough,” Hydrocarbon Processing, October 2001.

THE AUTHOR
	Tek Sutikno, PE, is a lead process engineer at Fluor Corporation in Houston. He has been responsible for a number of projects mainly in oil and gas production and processing. Dr. Sutikno has authored more than 10 publications, including patent applications in the areas of petroleum refining, gas processing and power generation. Dr. Sutikno holds BS, MS and D. Engr. degrees in chemical engineering, and an MBA degree, all from the University of Kansas.