Significant changes in the power industry over the last 15 years have created new challenges for owners and operators of today’s generation facilities. As market conditions fluctuate, those facilities must respond to the needs of the regional power markets they serve, the demands of local grid operators’ requests for power dispatch, and their own operating and business objectives. As operators navigate these conditions and work to meet their objectives, a significant impact has occurred on generating facilities.
The major reason for this impact is the reality that most plants in North America were not designed to be operated in a cycling manner. But major equipment within plants—from high pressure steam drums to control valves—receive the brunt of what happens during this mode of operation.
This article focuses on the negative effects changes have had on the control valves associated with the feedwater system of a typical power plant and gives several examples of common errors or improper usage of these valves. It also addresses what asset management practices can be implemented to mitigate or negate these effects entirely.
Before deciding to cycle a fossil plant based on economics, plant and fleet management have to consider many short-term and long-term effects. For example, along with damage high cycling can do to major components, control valves can exhibit signs of misuse from incorrect operation.
Every power plant has several critical or severe service control valves, and it is essential these valves be operated and maintained properly for the facility to run at optimum levels. This is true whether starting up, maintaining minimum load, throttling down from maximum capacity to bring renewables online or for other modes that require good control performance.
Successful application of any severe service control valve depends on these key attributes:
- Accurate design conditions
- Proper trim selection
- Proper installation
- Appropriate control strategies
- Proper maintenance
Understanding these attributes is critical to valve longevity and plant reliability whenever maintenance activities are planned for a severe service valve. This is because not understanding can cause significant problems for plant availability and uptime, which translates directly to plant profitability.
To illustrate the first key attribute of successful severe service valve application, let’s focus on the feedwater startup and regulating valves in a typical power plant (Figure 1).
The valves that are part of the boiler feedwater system endure some of the most severe operating conditions in a power plant. Before addressing key attributes for successful severe service valve application, a basic review of these valves and their purposes is helpful:
Although some power plants are configured with only one feedwater valve to handle both startup and normal operation, many plants use a valve pair that splits startup and full load operation duties. In a plant configuration that uses two feedwater valves, the feedwater startup valve is used during plant startup to send feedwater to the boiler (or heat recovery steam generator—HRSG—in the case of a combined cycle plant). During startup situations, there may be a significant pressure drop across the valve due to high boiler feed pump discharge pressure and very low boiler pressure, which requires installation of anti-
cavitation trim. Once adequate pressure is reached in the steam drum, the regulating valve should be opened to allow a sufficient amount of flow so that this valve is not operating at or below its minimum throttling point. From this point, the regulating valve should be controlling all the feed-pump output that goes through the boiler.
When the key attributes for successful application of severe service valves are not followed, the valve asset generally suffers from reduced longevity. Common issues seen in feedwater control valves are:
- Insufficient information to determine valve selection
- Poor control strategy
- Oversized valves
- Failure to specify tight shutoff
- Improper operation
- Entrained particulate
- Improper installation
A closer look at these issues as they relate to feedwater valves follows.
Issue #1. Insufficient information to determine valve selection: To ensure that final control elements continue to provide acceptable service and longevity, the conditions the valves face as a plant goes from startup to full load, through hot or warm restarts and any outlier conditions must be carefully considered. Many times the control valve initially sized and selected may meet original design and operating conditions defined on the data sheet but not necessarily be the most appropriate solution for every plant operation. Because of this, whenever a critical or severe service valve requires maintenance that potentially involves trim replacement or repair, these steps should be taken to replace in-kind maintenance:
- Review the actual process conditions by:
- Retrieving historian data from cold start to full load.
- Finding boiler feed-pump header pressure, temperature, flow (valve inlet)
- Finding downstream pressure (valve outlet) for:
- D/A, condenser or LP drum for BFP recirc
- Economizer inlet pressure for FW regulator
The goal of these steps is to define a comprehensive set of service conditions that represent the actual process the valve will experience rather than replacing the same valve trim originally designed, sized and selected based on the engineering datasheet. That data may no longer be appropriate to support the same trim selection.
Issue #2. Poor control strategy: Although a valve may have been sized and selected properly based on accurate and complete operating conditions, it may still experience reduced longevity if it isn’t controlled properly by the operator and/or the distributed control system.
A primary example of improper control strategy is an incorrect transition between the startup and regulating valve. To maintain stable drum level control, this transition should be done at the point where the startup valve is at about 80% of its capacity and the regulating valve is at about 20% of its capacity. Implementation of this transition in the control logic has proven to be a successful practice for avoiding control issues between the two valves that are many times simultaneously throttling the output of the boiler feed pump. This practice is often referred to as the “80/20 rule.” Significant deviation from this rule results in unstable drum level control and operation of the feedwater control valves in a range that reduces the service life of the valve assets.
One critical feature of this rule is that it avoids operation of the regulating valve below the minimum throttling point, which is defined as the minimum point of operation for effective throttling (and thus control) for the control valve. Also, depending on the specific design, the valve trim may not be able to effectively stage the pressure drop to prevent cavitation if throttled below this point.
Operation below the minimum throttling point typically leads to erosion of the seating surfaces and to increased frequency of valve trim replacement. Since many combined cycle plants had been in a cycling mode with upwards of 250 startups and shutdowns per year, the damage inflicted to these valves appears daunting. If the “80/20 rule” can’t be met exactly, it should be met as closely as possible. Figures 2 and 3 show the damage that can result from operation at or below this minimum throttling level for appreciable amounts of time. This damage prevents proper plug and seat contact, resulting in decreased shutoff capabilities for the valve assembly. The decreased shutoff of the valve assembly can then lead to further and accelerated trim erosion damage.
Both symptoms lead to further leakage across the seating surface. To mitigate the damage from a poor control strategy, implementation of the “80/20 rule” should be considered. This type of control strategy can greatly increase the longevity of both these valves and lead to reduced maintenance and replacement part costs. Although 10% has been mentioned as a minimum throttling point, which is a conservative travel value to consider as the lower limit, specific minimum throttling values are available from valve manufacturers for each trim set so that control logic, including low travel cutoffs, can be customized for the given valve construction.
Issue #3. Oversized valve trim: Feedwater valves tend to be oversized for an application. Most times this is because a design accommodates an operating condition occurring during a plant “trip” when the safety valves lift and sufficient flow through the boiler or HRSG is critical to maintain. The valve pressure drop is designed to be minimal for this, many times only 20-30 psig.
While accounting for “maximum” condition ensures the valve will have sufficient capacity should this rare condition occur, it also causes the valves to operate at 30-40% open during normal operation. This means that during minimum conditions, the valve is most likely operating at or below its minimum throttling point. As mentioned above; lower lifts expose the seating surfaces to premature erosion during startup. As mentioned in Issue #1, the proper steps to ensure this doesn’t continue involve developing a comprehensive set of operating conditions that fully define the actual operating range the valve ought to be controlling.
Issue #4. Failure to specify tight shutoff: One of the more surprising discoveries about currently installed and operating valves in feedwater service is that they were initially specified with poor shutoff. Many times a Class II, III or IV shutoff is specified, which allow appreciable flow across the valve seating surface when the valve is in the fully closed position. As a best practice, all valves in feedwater service should be Class V to prevent erosion from leakage flow. Leakage classes of Class IV or lower provide enough flow to rapidly erode valve seats. The chart in Figure 4 illustrates the significant difference in the various shutoff classes for control valves.
Excess leakage also can cause excursions in drum level, which can cause excessive blowdown to maintain level and excessive pump start/stops if level cannot be maintained with blowdown. This can lead to pump motor damage.
Class V shutoff using 800 pounds per lineal inch of seat circumference of actuator seat force should be specified. Class V is a water test at service pressure where all other tests are at 50 psi air. Existing actuator sizing should be validated, and an actuator change may be required. Valve diagnostics will insure actual shutoff force meets required force, and valve travel cutoffs (which should be set within the digital valve positioner) will insure the positioner saturates to full output and that full seat load is available, while also ensuring the valve does not try to control below the minimum throttling level.
Issue #5. Improper operation: An example of improper operation is when an operator decides to avoid using the boiler feedwater startup valve (Valve #1 in Figure 1) altogether because of personal preference or operation methodology. Choosing to start the plant and fill the downstream steam drum using only the boiler feedwater regulating valve (Valve #2 in Figure 1, also commonly referred to as the drum level valve) leads to cavitation damage, which further leads to leakage and subsequent high pressure feedwater cutting the seating surfaces of the valve trim.
In general this leads to reduced life of the final control element. If not repaired correctly, the result may be an inability to properly control plant startup or maintain steady state operating conditions, including critical drum level set points, when future plant operation demands it. Although undesired plant operation from the human element is never completely reduced, it is critical that plant operators understand the proper operation of the feedwater startup and regulating control valve pair. Since the startup valve’s purpose is to handle the high-pressure drop conditions, placing the burden of handling those conditions onto the larger regulating valve (which isn’t designed to prevent cavitation or to throttle at low Cv values) serves to reduce the longevity and reliability of that asset.
Issue #6. Entrained particulate: The issue of entrained particulate becomes more urgent when small orifice style trims are used (mainly in designs intended to prevent cavitation). Problems from flow-accelerated corrosion and the presence of magnetite manifest themselves in these types of trims by plugging the flow holes that are required to properly stage the pressure drop to prevent cavitation and its damaging effects to the valve trim.
Although small orifice trim styles are very effective when the feedwater is clean, the presence of particulate quickly degrades the performance and service life of this type of trim. If this is experienced, an upgrade that uses a dirty service style of trim designed to properly stage the pressure drop and also allow for particulate passage should be considered. These trim types incorporate an inherent protected seating surface so that as the valve plug moves from the closed position to a certain set point, high-pressure feedwater flow has sufficient area to pass through so that localized high velocities across critical seating surfaces can be avoided.
Issue #7. Proper installation: Startup and commission of the entire power plant is a critical step in the construction process of a power plant. It is important to ensure all piping and related components are suitable for operating pressures, that all welds were performed correctly and the joints don’t leak, and that all debris in the piping system has been properly flushed so that sensitive equipment is not damaged.
Many times, proper flushing of the feedwater lines is not performed, and valves with small orifices as part of the design (for example, with small holes in the cage) are plugged and damaged from debris caught between the plug and cage, causing scoring and galling of critical valve trim surfaces. A highly recommended best practice is to use proper flushing trims available from the valve manufacturer to ensure that these assets are not compromised even before they are put into service.
By focusing on analyzing existing maintenance practices and making a step toward predictive and preventative practices, unscheduled plant downtime (equivalent forced outage rate) can be minimized while ensuring assets operate at optimum levels. Aside from the significant benefits of implementing these practices, improvements in reactive maintenance methodology can also contribute greatly to asset optimization and ensure that during outages (planned or unplanned), the correct maintenance is done. Through analysis and enhancement of existing reactive maintenance practices, greater levels of plant performance, availability and reliability can be realized.