The 7F Users Group and CCJ are working together to expand the sharing of best practices and lessons learned among owner/operators of large frame engines. One of the user organization’s objectives is to help members better operate and maintain their plants, and a proactive best practices program supports this goal. Consider submitting to the 2022 Best Practices Awards Program: Deadline April 1.
Below are some 7F entries in the program receiving Best Practices Awards. They speak to work done by your colleagues at 13 simple- and combined-cycle generating plants. The more than two-dozen best practices are likely to offer one or more ideas for improving safety and performance at your facility. Consider attending the upcoming 7F Users Group in Dallas: May 23-27.
Barney Davis Energy Center
KPI index for water chemistry
Barney Davis Energy Center, Corpus Christi, Tex, implemented an EPRI-recommended water-chemistry KPI (key performance indicator) index. The plant’s original analyzers were updated and more instruments were installed, reported Plant Manager Jay Langley. Two dissolved-oxygen analyzers were added, plus one sodium analyzer; four silica analyzers were replaced.
All of the new analyzer readings were added to the DCS. KPI-index action alarms, Levels 1 and 2, also were added to the DCS in accordance with EPRI recommendations. An action plan was developed to guide operators when an alarm level is reached.
When conducting water-chemistry tests with manual grab samples, the number of samples and tests performed increased. Grab-sample data added plant-rounds documentation which is transferred to the data historian using a hand-held device. Next, the data are sent to an Excel file where the KPI scoring system is applied. It gives the plant an indication of where it needs to go regarding water chemistry. The score sheets are archived to allow quick reference to previous scores and to identify developing trends.
Replace haz-gas detector without compartment entry
Goal was to make replacement of turbine-compartment hazardous-gas detectors safer by eliminating the need for personnel entry into the compartment to perform this task. The
best practice shared facilitates detector replacement, which is done from the roof and can be performed in a few minutes—during an overnight shutdown, for example, without impacting unit availability. Following completion of site LOTO and safety procedures, the rail that contains the bad detector can be unbolted by removing four small bolts.
Change-out of a haz-gas detector went from 24 hours to 20 minutes. Cost of the procedure dropped by several thousand dollars because scaffolding is no longer required for access. Employee safety also improved because staff is not exposed to the high compartment temperature.
Bastrop Energy Center
Protect plant personnel, visitors with machine guards
Safety best practice: Install machine guards to protect operators and others against injuries caused by exposed rotating shafts. They are your first line of defense. Bastrop Energy Center, Cedar Creek, Tex, identified a potential hazard with its diesel- and motor-driven fire pumps. Plant Manager Kelly Fleetwood reported that the shafts on both sides of the pumps were exposed, creating a potential hazard.
Plant personnel fabricated the protective covers for the exposed areas. Expanded metal guards were used to allow visual inspection of the packing follower and to monitor bearing lubrication.
Calhoun Power Co
Objective of Plant Manager Mike Carter is to reinforce the safety culture at Calhoun Power Co, a 4 × 0 dual-fuel plant in Eastaboga, Ala, by making ongoing enhancements to plant systems that protect site personnel and visitors. Two safety enhancements recently implemented are these:
- Eliminate need for entry into a confined space to pump out the contents of a drain tank. The simple solution: Extend the tank’s drain line so staff can connect to the vacuum truck without entering the tank well.
- Eliminate need for a ladder by installing a permanent safety platform to reach over closed-cooling-water pipes to access the generator belly-pan level switch requiring a monthly test.
Central Eléctrica Pesquería (CEP)
Benefits of lean-outage planning, execution
Following this Mexican plant’s first hot-gas-path (HGP) outage in 2018, personnel, under the direction of Plant Manager Mario Ontiveros, met to discuss the pain points identified during the work just completed. Goal was to reduce the time needed to conduct scheduled outages and to optimize resources by ensuring the integrity of people and equipment. Savings would be near immediate because there were two more gas turbines at the 7FA.05-powered 3 × 1 combined cycle in the HGP queue.
Among the many pain points identified were these:
- Scaffolding installation.
- Removal of blades and nozzles.
- Lifting and dismantling of the package roof.
The methodology selected was “Lean,” which focuses on quality assurance, time reduction, resource optimization, and effective communication. The rigor applied to outage planning and execution was evident from the details provided in the best practice—including the following:
- A visual scaffolding installation plan was implemented and strategic scaffolds were assembled prior to turbine disassembly.
- The scaffolding plan allowed work in tandem in the disassembly of turbine elements and combustion chambers.
- A lifting plan was implemented to disassemble the entire roof, previously removed in sections. Plus, pre-location of mechanical lifting elements for easy availability.
Results: The execution time of the second and third HGPs was reduced by 15% compared to the first outage, which lasted 19 days. Quality standards were maintained and there were no problems during and after the outages. Finally, the OEM recognized CEP’s outage team as the best in Latin America based on the results of the second and third HGPs.
Optimizing the steam-turbine cooling process
Challenge: Develop a methodology to cool the steam turbine in the shortest time possible to conduct minor maintenance while adhering to the manufacturer’s cooling curves and delivering the output required to satisfy contractual obligations. Plant had determined previously that this could be accomplished by operating one of the three gas turbines baseload at 210 MW and the other two in temperature matching for nine to 12 hours.
Plant and OEM personnel determined that, in this case, dry-air cooling was the optimal method for rapid turbine cooldown. It did not adversely impact the cyclic life of the rotor or the critical path of the outage. The method consists of injecting a regulated amount of air gradually into the HP and IP stages of the turbine while assuring that the temperature differentials between the stages were not so great as to stress the rotor. Goal: Cool the unit to the point where the turning gear could be shut down.
About 30 hours of turbine operation were gained using this procedure in place of the previous (standard) method.
Redesign of sulfuric-acid dosing system eliminates leaks
Pesqueria’s sulfuric-acid dosing system for water treatment typically experienced three or four leaks per month totaling up to about 5 liters of acid—jeopardizing the integrity of personnel and equipment. A team was formed to address the root causes of the problem and correct them.
Steps taken included the following:
- Changed all the PVC suction and discharge hoses for the diaphragm pumps to chemically resistant PVDF pipe.
- Upgraded pipe to Schedule 80 from Schedule 10.
- Installed a strainer filter with blocking valves to prevent fouling of pump check valves, thereby reducing the exposure of personnel to the risk of intervention.
No acid leakage event had been identified at the time this best practice was developed, about 1500 hours after the action plan was implemented.
Effingham County Power
Maintaining fire protection during an outage
Effingham County Power’s fire protection system consists of one 2000-gpm motor-driven pump, one 2000-gpm diesel-driven pump, one 25-gpm electric jockey pump, and associated equipment and controls. The jockey pump maintains system pressure above the start setpoints for both fire pumps during no-flow conditions. Pressure switches automatically start the pumps when system pressure drops below the setpoints.
Power for the electric pumps and their associated equipment and controls is supplied from a common 480-Vac breaker housed in a switchgear with station service its only source of electricity. When the station-service transformer is taken out of service for maintenance, the two electric pumps are de-energized. Because the jockey pump is de-energized, it is unable to maintain system pressure above the starting setpoints of the fire pumps.
With the electric pump de-energized, the diesel unit will start once system pressure drops below its associated pressure-switch setpoint. When the diesel pump starts, it activates the plant-wide fire alarm; all staff and contractors must stop work and evacuate the plant when the alarm sounds.
Plant management at the 2 × 1 combined cycle in Rincon, Ga, headed by GM Bob Kulbacki, needed a way to keep the fire main pressurized when the plant was on backup power and the electric fire pumps were de-energized.
Plant personnel ran the diesel pump intermittently to maintain system pressure, to prevent erroneous fire alarms and work stoppage. This resulted in unbudgeted diesel fuel costs and excessive emissions, which could possibly lead to exceeding air-permit limits.
Staff identified the nearest motor control center supplied by a backup source of power when station service was de-energized. The closest spare breaker was about 150 ft away and required extensive trenching to run supply cables to. But a welding receptacle, which has a secondary source of power, was located only 50 ft from the fire-pump building. Staff determined that the rating of the welding supply power was sufficient to operate the jockey pump.
The plant purchased a double-throw safety switch, one plug, and the necessary cable to connect the welding receptacle to the switch located at the fire-pump building. Total cost of the materials to complete the project was $500. The two plugs were connected to the supply cable to make a jumper cable to connect the welding receptacle to the fire-pump building.
During an outage, with station-service power de-energized, the switch was permanently wired to the jockey pump’s primary power supply and one end of the temporary supply cable. Next, the temporary supply cable was connected to the welding receptacle. Then the welding receptacle was energized and staff swapped the jockey pump from normal to temporary power.
The jumper cable is stored in the fire-pump building to ensure its availability at all times. Prior to securing station-service power, the jumper cable is connected to the designated welding receptacle. When station-service power is de-energized a technician energizes the welding receptacle and swaps the safety switch from normal to temporary power, thereby providing power to the jockey pump. Fire-protection integrity is maintained in this configuration because the diesel pump will start if system pressure drops below the setpoint.
With the jockey pump in service, spurious alarms and work stoppage have been eliminated. This project has been a cost savings because outage interruptions have been reduced and the need to operate the diesel pump intermittently to maintain system pressure is no longer required.
Assuring voltage-schedule compliance
The transmission operator has provided Effingham County Power four voltage schedules which change throughout the day: 0000-0600, 0601-1800, 1801-2100, and 2101-2400. Control room operators are required to maintain the schedules within a ±2 kV band. In the past, the CRO ensured the plant was in compliance with the schedule by visually confirming the plant’s “white line voltage” and adjusting the generator’s output voltages accordingly.
To avoid operating outside the control band, a generic alarm was developed to alert operators. A narrower band was established but did not vary with the changes in the voltage schedule.
To prove to the regulatory agency that the plant was in compliance, staff developed a monthly comparison spreadsheet showing plant voltage and the allowable voltage band for each minute. This required manually inputting over 44,000 data points into the spreadsheet and reviewing to verify the information was correct.
The plant’s output voltage values were populated into the spreadsheet from the DCS historian. Obtaining the necessary data for the control bands required review the shift turnovers for the past month and manually inputting these values. Once all data were entered, the spreadsheet was reviewed for compliance and saved for future audits.
The voltage schedule is determined by Georgia System Operations Corp under NERC standard VAR-002 and issued to the plant daily. If a change is made to the voltage schedule, the alarm will not change based on the current logic. This increased the potential for human error in keeping the plant operational within the established schedules.
To alleviate this issue, staff created a series of logics and graphics in the DCS (Emerson Ovation) that would allow the operators to select which schedule is currently in effect. This input is then compared to the time of day to generate the correct alarm band for the plant. The programmed alarm will alert the CRO if the plant’s white line voltage deviates from the designated voltage schedule.
If the operator continues to operate outside of the required voltage schedule, a second alarm is
generated notifying the CRO that the plant is still operating outside of the voltage range and that it is approaching compliance limitations.
The voltage schedule, control band, and time are new logic points designated in the DCS system. All values are fed into the DCS historian for retrieval as needed to show evidence of compliance to regulatory agencies. The need for manual data input has been eliminated, reducing human-error issues and saving time during monthly reviews of VAR-002 data.
New graphics and logic have allowed both operators and management to ensure the correct voltage schedule is maintained. Additionally, the alarm points and trending capabilities have enabled staff to accurately determine if the voltage schedule is consistently maintained for reporting requirements.
The monthly spreadsheet can be updated quickly and the plan can show compliance in the time it takes to retrieve the data from the historian.
Relocating grease fittings simplifies PM
The original design of the inlet-chiller cooling-tower fan specified the location of grease fittings inside the fan enclosure, limiting accessibility to the fan bearings for maintenance. To perform bearing PM, technicians had to place a LOTO on each fan and post a confined-space permit before entering the enclosure to grease the bearings. The prerequisite steps to establish a safe work environment took approximately two hours for each fan, while greasing the bearings took only 10 minutes.
Greasing of fans typically occurred when the chiller was shut down, causing no loss of generation. But there were several instances when a fan required greasing when the chiller was online. In these cases plant output was reduced by about 2 to 3 MW.
Technicians were tasked with finding a more efficient and safe method for performing this PM.
The best option discussed was to locate the grease lines on the outside of the fan enclosure. This would eliminate the need to secure the fan and issue a confined-space permit. Staff purchased the ¼-in. stainless-steel tubing and fittings required for about $500. The remote fittings are attached to the bearings using the ¼-in. lines; flexible hose supplies grease to the fan housing platforms. A hole was made in the metal housing, and a zirc fitting mounted on the housing, to allow remote greasing.
Split bearing facilitates fan bearing replacement
One of Effingham’s chiller cooling-tower fan bearings failed, resulting in the loss of tower cooling efficiency and of plant output. Because the fans are located 42 ft above grade, and the fan assembly is about 8 ft across, replacing the bearings without use of a crane would not be safe, or technically possible. With a crane it would take two days to remove the fan, replace the bearings, and reinstall the fan—with an out-of-pocket cost of about $3200.
An easier and less expensive way to replace the bearings was needed. Staff researched alternative solutions for replacing the OEM’s single-piece bearings and opted for using slightly more expensive split bearings of the same rating, which would not require fan removal. Another benefit of the split design: Less time to replace a bearing by a factor of four—4 hours versus 16 for the original.
The initial installation of a split bearing was time-consuming, and the limited work area made it difficult. One reason: The mounting holes for the split bearings were not a direct match to those for the OEM’s bearings, so the mounting base required modification. Additional attention also was required to maintain correct shaft alignment to ensure proper fan operation.
Crews were rotated during the installation to keep everyone fresh and working safely. The new bearings were installed without incident and the plant has had no issues since the fan returned to service. Over time, as OEM bearings fail from wear and tear, they will be replaced with split bearings.
Squeezing a new air compressor into existing space without a forklift
A compressor failure left Effingham County Power with only one source of compressed air. If the remaining compressor were to fail, plant production would be lost. A rental compressor was brought in to mitigate risk. However, the rental did not have auto-start capability, so it was unclear if this unit could be brought online fast enough to prevent a loss of control air, without which the plant would be forced to shut down.
A new compressor was ordered and quotes were received to remove the old unit and install the new one. Quoted cost for removal and installation was more than $11,000. The maintenance crew was challenged to find a more cost-effective and safe method to replace the plant air compressor.
The air-compressor room configuration made it unfeasible to remove a wall, or the roof, to make the switch. The size and weight of the compressor was a consideration when planning the replacement. The limited space in the compressor room did not allow for the use of a forklift or other powered equipment. With building modifications unacceptable, team members decided to roll out the old compressor and roll in the new one using caster rollers and jacks.
The old compressor was lifted using the bottle jacks and placed on channel iron supported by the four swivel caster rollers. This allowed maneuvering the compressor in the restricted space. The old unit then was rolled out of the compressor room and the process reversed to install the new compressor. The exchange of air compressors was completed safely in about 32 man-hours at a cost of $1100.
Ensuring quality service by OEMs during gas-turbine outages
How does a powerplant manager ensure the OEM delivers quality work during turbine/generator outages? Here “quality” is defined as work safely completed with no rework required. Even with a strong contractual services agreement, an OEM’s failure to deliver the job correctly the first time can be a lose/lose situation for both the owner and OEM.
Powerplants may no longer have the talent available to adequately ensure OEMs perform their work properly. In addition, the quality of the OEM’s talent, based on Elwood Energy’s recent experience, appears to be in steady decline with crews being thoroughly fatigued towards the end of outage season.
Elwood Energy, a nine-unit simple-cycle plant, tackled this problem by contracting Viking Turbine Services to provide oversight, typically on day and night shifts, through major turbine and generator work. With major work being performed on one or more turbine/generators, Viking maintained a skilled and dedicated eye on the OEM’s performance. This freed-up Plant Manager Joseph Wood’s small staff to administer safety programs, perform other outage-related work, and respond to dispatches of other units.
The following is one example of the third-party service provider’s value: While doing routine checks, the company’s personnel noticed that two of the transition pieces completed on day shift had significant gaps between the bullhorns and bullhorn support blocks. You’re likely aware that bullhorns must be snugged down tight prior to recording set-back clearances.
Viking requested that the OEM’s contractor tighten down the bullhorns and recheck set-backs. Proper fit-up remained elusive for some of the transition pieces and the OEM decided to change all bullhorn blocks.
Over the last 10 years and numerous turbine outages, Elwood has experienced two outages requiring significant rework—both when the plant did not have third-party support. Numerous other outages have been performed with oversight that have not required rework. During these outages, Viking routinely identified issues during routine observations and during agreed-to quality hold points.
Essential Power Newington
Rehabilitation of EHC fluid boosts starting reliability
During a routine plant cycle/startup, the right-side HP steam control valve failed to open and allow steam to flow to Essential Power Newington’s D11 turbine. The startup was aborted, the electrohydraulic (EHC) control system was secured by LOTO, and the servo valve controlling fluid flow to the valve actuator was replaced. Newington was restarted and the HP control valve responded as expected; plant operation was restored to “normal.”
When the failed servo was shipped to Moog for rebuild, an in-depth failure analysis found what appeared to be a varnish-like substance on the nozzle top and flapper to the torque motor. An EHC oil sample was sent to a certified laboratory for contaminant (varnish, water, metals) analysis, an estimate of the oil’s remaining service life, etc.
Lab results: The deposits, originally thought to be varnish, actually were carbon. At the time of the servo failure, the EHC system included a kidney-loop varnish removal/filtration system. Site personnel, who relied on the fluid OEM for sample analysis (no charge), came to learn that the lab effort did not fully check the fluid for all required parameters. The only parameter that had dropped recently was resistivity and the reason for that was to be investigated.
Staff’s first thought was to dump the oil and flush the system, which would have been extremely expensive.
Plant personnel, led by Plant Manager Tom Fallon, embarked on a multi-faceted plan to review contaminated-oil sample results, discuss possible solutions with industry experts, implement corrective actions to mitigate system condition, and avoid a fluid change-out.
Newington staff worked with Advanced Fluid Systems, a fluid-power solutions provider, and filter OEM Hy-Pro Filtration to develop the following approach for removing contaminants from the system:
- Change the EHC fluid filter media to eliminate sparking.
- Inject dry instrument air into the head space of the tank.
- Rent an electrostatic contamination removal skid to pull out the carbon deposits.
- Use improved oil analysis to monitor trends.
Also, a hydraulic fluid pump rep visited Newington to perform a system walkdown and make sure there wasn’t some abnormality causing the fluid contamination.
Following the EHC system changes, staff performed semi-monthly sample analyses to monitor trends in fluid cleanliness. With the results of the analyses showing improved fluid characteristics over time and the visual indication of the fluid becoming more transparent, staff knew the plant’s approach was working.
By discussing Newington’s problem with many industry experts, staff was able to get several opinions on what to look for regarding the source of contamination, which ultimately was carbon buildup. With expert input and the plant’s commitment to finding a solution, staff was able to effectively clean-up the system and eliminate the need to replace the existing fluid.
Lesson learned: By focusing on an issue and not letting the path of least resistance become your answer, you can effectively eliminate a problem long-term. Had staff simply changed the fluid without performing the other steps in the process, carbon contamination would have reappeared in a matter of time.
PI tools help improve situational awareness, work processes
This best practice is divided into two different, but similar, business-challenge paths and targeted solutions surrounding the use of existing PI tools and control-room recordkeeping. Both PI tools described below have proven extremely successful at Newington. Each has its unique success stories, but both ultimately provide excellent efficiency and awareness improvements to the site. Improving facility situational awareness and focus while reducing the many inherent distractions, and otherwise manual processes, is always a benefit.
No. 1. Situational awareness and information-sharing is a best practice that should be used across multiple fronts. With regard to operational awareness and NERC compliance, Newington decided to further utilize the PI Notification Tool to automatically send e-mails to communicate several different key plant operational and NERC compliance conditions. PI Notifications included chemical-tank levels, breaker operational counters, key equipment temperature monitoring, facility forced power oscillations, and NERC VAR-002 notifications.
Situational awareness for facility personnel is crucial to ensure timely and accurate response to changes in plant conditions. Understanding trends and future consequences are easily tracked using PI Notification tools at the facility level. These notifications are used for real-time monitoring, e-mail alerts, compliance requirements, and chemical-inventory reorder processing.
Newington personnel now are better tuned-in to operational and equipment status—including feed-pump bearing trends, voltage schedule, and Automatic Voltage Regulation (AVR)/Power System Stabilizer (PSS) status as required by VAR-002 requirements, chemical inventories for automatic notifications to chemical suppliers, breaker motor and cycle counters for equipment monitoring and forced megawatt oscillations. This allows multiple key personnel to be aware of conditions in the facility and increases the efficiency of many otherwise manual tasks.
Staff created all PI Notifications based upon input from the operations manager on needed awareness tools. Several different notification categories were created, as noted above. Notification and alert levels were set up to track initial concerns and then, if necessary, any communication requirements based upon the condition.
Currently, 10 notification elements are used by PI. Some look at real-time statistics, others at trends and future objectives necessary for compliance and equipment protection. To date, several notification e-mails have been used. They provide excellent awareness to plant conditions. Daily automated e-mails to the facility’s boiler chemical provider from PI allow for automatic reorder tracking and awareness, eliminating the need for phone calls or e-mails to order chemicals
No. 2. CRO distractions are numerous and having the ability to reduce them is always a challenge. During the course of any major operational evolution—such as plant startup or shutdown—the ability to focus on the task at hand is extremely important. One of the requirements in any control room is log-keeping. For several years, the facility has used an electronic program called eLogger for this purpose, but entries in the database still required manual input.
In an effort to further reduce operator distractions, a new eLogger log-entry automation tool was developed, tested, and implemented by staff. The PI database was interfaced to the e-Logger application, providing PI-generated log entries for key plant and equipment status changes, greatly reducing the amount of manual log entry required by employees.
Automating logging events is an extremely effective solution for removing distractions, yet still accomplish the record keeping needs of the facility. Automatic logging was setup for conditions such as blower or pump starts/stops, turbine starts/stops, etc.
Using PI to communicate automatically with the eLogger database allows for a reduction in the hundreds of otherwise distracting manual logging tasks required by the operations staff. This drives better focus on the control system for operations to perform critical tasks while still completing certain required log-keeping “behind the scenes.”
A plant employee designed and constructed the application necessary to allow the PI and eLogger databases to communicate continuously. The application monitors numerous PI tags for state changes then passes the event to eLogger, where a log entry is automatically generated.
There are currently over 60 different PI tags communicating with eLogger and recording as necessary. This list will be expanded continually based on need and efficiencies following input from others. Improved focus and attention on the control boards has been accomplished with reduced manual logging needs.
Hunterstown Generating Station
Attention to leaking valves slashes water use
Goal to reduce cycle steam and water losses focused on identifying leaking valves and making repairs. A vendor was contracted to perform valve leak surveys beginning in late 2016. From 2017 to 2019, many valves were tested and numerous large-to-medium leaks were documented. Valve repairs and replacements implemented during outages were highly successful. Cost of demin water production dropped by $100,000 over two years while capacity factor increased by 8.3% during the same period.
NDT program eliminates forced outages caused by steam-drain failures
In 2016 and 2017, Hunterstown, a 3 × 1 combined cycle under the direction of Plant Manager Tom Hart, experienced eight unplanned outages to repair drain-line leaks caused by steam erosion. These outages represented 18 days of unavailability in the two-year period. The leaks occurred most often during startup and shutdown operations and presented potential hazardous situations to personnel because of their locations.
The plan developed to eliminate these forced outages included the following actions:
- Review work-order and weld-log histories to identify the most frequent recurring failure locations—HP and IP continuous blowdown drain lines. These lines are characterized by large differential pressure gradients and two-phase flows which accelerate pipe-wall erosion rates.
- Create prioritized inspection plans for HP and IP continuous-blowdown drain elbows. Include drain-line isometric drawings showing elbow IDs based on upstream MOV tags. Elbow IDs were sequentially numbered in the direction of steam flow to more easily track x-ray inspection images and reports, and to build a tracking spreadsheet for repairs.
- Radiograph small-bore drain elbows during normal plant operation without removing insulation or lagging, using large receptor panels to capture the entire elbow and part of the downstream piping in single-exposure shots.
Inspection and repairs eliminated unplanned outages attributed to HP and IP steam-drain-line fitting failures. A rope alternative to scaffolding for access to failure locations saved tens of thousands of dollars. The bottom line: There was no spending for scaffolding and insulation removal.
Maximize GT generation with evap coolers in service
Maximize use of gas-turbine evaporative coolers by implementing control logic improvements to automate starting and stopping of the evap coolers based on the following: ambient and inlet air temperatures, IGV angle, inlet bleed heat (on or off), and evap sump level. The cost of implementation was about $16,000 for a first-year gross-margin increase of approximately $25,000.
Benefits of transitioning from CO2 bottles to bulk storage
When generator purges were needed, Hunterstown relied on portable 160-liter CO2 bottles. Pressure in the bottles decreased over time (outdoor storage, exposed to sunlight), rendering the bottles minimally effective. Multiple bottles were needed to complete a single purge.
Plus, time-consuming operator manipulation of the bottles was needed to complete each purge. Staff had to transition between units and the CO2 storage facility and to secure spent bottles and line up new ones. Such maneuvering was critical during emergency generator purge situations because of the time required to begin flowing CO2 to a generator. On more than one occasion, nearly 30 minutes elapsed between the time a purge was started and gas began to flow.
Solution was simple: Install a bulk storage system with a capacity of 1550 gal having these features:
- Electric vaporizer to convert liquid CO2 to vapor to prevent two-phase purge flow.
- CO2 condenser to convert excess vapor back to liquid, thereby preventing tank losses to atmosphere.
- Safety shutoff device to ensure liquid CO2 cannot reach the generators.
- Instrumentation for tank level and pressure with data routed to the DCS to monitor consumption rate and to reorder CO2 when necessary.
Cost of the system, installed in February 2020, was less than $350,000—including equipment, engineering, installation, and commissioning. Bulk storage eliminates bottle demurrage charges of about $11,000 annually and simplifies operator actions to start CO2 flow, saving several minutes over the bottle alternative.
Expectation is that the four units on site can be purged twice before a refill is needed, while retaining sufficient gas for two more unit purges (total of 10).
Marcus Hook Energy Center
Upgraded stack for cooling-tower fans facilitates maintenance
Marcus Hook Energy Center, a 3 × 1 combined cycle under the direction of Plant Manager Frank Meade, has a 12-cell mechanical-draft counterflow cooling tower with multi-speed fans. Blade-tip-to-stack contact was causing wear on both components. The original stacks were made of molded fiberglass, with only vertical ribbing. They were constructed in sections, through-bolted together on the inside. If hardware loosened while in operation, a confined-space entry and extensive scaffolding would be required to replace or tighten.
The new stacks have ribbing in both the vertical and horizontal directions, making this design less susceptible to flexing and reducing the probability of blade-tip-to-stack contact. This also lessens the chance of hardware coming loose. The new stacks connect sections with a completely exterior face-to-face flange, allowing the tightening of all hardware from the outside—thereby contributing to safer, less-expensive, and more-effective routine maintenance.
The increased fan-stack rigidity allows use of gear reducers with standard output shafts—typically more readily available and less costly than the original gear reducers. This allows fan blades to be located lower in the stack—hence closer to the sidewall—increasing cooling efficiency by decreasing tip vortex. The plant has replaced several of the OEM stacks with the improved design and will continue to do so over the next few years until all cells are retrofitted with the new stacks.
MEAG Wansley Unit 9
Bearing-tunnel fire alarms enhance personnel protection
Plant Manager Timothy Williams, MEAG Unit 9, Franklin, Ga, wanted to assure a high level of safety for personnel working inside the gas-turbine exhaust enclosure. NFPA-12 provides guidance for carbon-dioxide extinguishing systems typically specified for generating plants powered by gas turbines. The standard provides the minimum requirements for a system designed to flood the compartment with CO2, which does not support combustion or life, should a fire be detected.
There is the potential that an individual could be working inside the enclosure, unaware of a fire event, and could be a long distance from the nearest exit. Example: What if the CO2 vented out of the adjacent lube-oil drains and into the exhaust enclosure?
At MEAG Unit 9, as in most plants, mandatory signage is in place as a warning, but is that sufficient? Staff decided an additional measure of protection was required since there were areas where CO2 could be present, but no visual or audible alarms were available. So, they installed additional three-horn-and-strobe-light combination alarms at the following locations:
- Near the bearing-compartment vent.
- At the entrance door to the exhaust section.
- Inside the exhaust-section enclosure.
This now provides personnel working inside exhaust enclosure immediate and effective notification of the danger in the unlikely event a fire were to occur in the compartment.
Training videos promote shop safety
Following an extensive maintenance-shop overhaul and retooling at Plant Rowan, Salisbury, NC, staff was challenged by Plant Manager Chris Lane, supervisors, and the site safety council thusly: How do we ensure everyone operates and maintains the equipment according to our company safety policy, common industry practices, and manufacturer equipment procedures?
Plant operators may only use a certain piece of equipment infrequently—such as during outages. The lack of daily use concerned staff because proficiency conceivably could be lost during periods of low maintenance.
The management team, together with the site safety council, provided direction for the creation and implementation of a series of Just-in-Time (JIT) “refresher” videos. A primary goal of the effort was to keep the videos short, each averaging only about five to six minutes per piece of equipment, but including as much information as possible.
The belief was that in this amount of time an experienced operator could demonstrate safe and effective operation or maintenance of a given piece of shop equipment with adherence to all company safety policy and equipment procedures. Equipment covered included pedestal bench grinders, horizontal band saw, drill press, various machining coolant systems, etc.
No narration was used; narrating the videos was deemed time-prohibitive at Plant Rowan. Instead, video editing software was used to add detailed on-screen “pop-ups” during final editing. This included pertinent information and hazards at each step of the equipment’s operation. Examples: When PPE is required and what type; tips for safe operation; hazards and examples of improper operational practices, etc.
Once completed, the videos were loaded on a 40-in. wall-mounted LCD monitor located in the maintenance shop for anyone to access prior to equipment use and as part of their pre-work job safety analysis.
The videos have been credited with contributing to significantly fewer occurrences of improper tool/equipment use—such as non-ferrous metal fouling on bench-grinder wheels, prematurely worn metal band-saw blades, and less wear and tear on equipment. Plus, there have been no safety issues.
Rathdrum Power LLC
Catwalk around clarifier promotes safe maintenance
Rathdrum Power, a 1 × 1 combined cycle located in Rathdrum, Idaho, managed by Richard Ihrig, is permitted as a zero-discharge (ZD) facility. As such, the plant does not discharge any process water. Undesirable solids are removed from the various process-water streams by mechanical means in the ZD area of the plant.
An issue with the plant’s ZD system was the design of its ageing clarifier. Sodium carbonate (soda ash) carries over and plugs discharge holes around the clarifier’s weir. Occasionally plant personnel have to clean the weir and service the sulfuric-acid discharge piping, tasks requiring use of an extension ladder. Putting up a ladder and climbing to the top of the clarifier to tie off the ladder posed safety concerns and was not viewed as efficient use of employee time.
With help from a local engineering firm, staff decided to resolve the problem this way: Build a catwalk around the outside of the clarifier so O&M personnel could access all areas at the top of the vessel. It was designed with a grating of Type-304 stainless steel and handrails and supports of carbon steel.
The catwalk enables access to the clarifier weir and sulfuric system without need for an extension ladder requiring fall protection. With easier access, cleaning intervals have been reduced. Plus, fewer personnel are required to do the work.
Woodbridge Energy Center
Rearrange collector air intake for better performance
The D11-A steam turbine at Woodbridge Energy Center, Keasbey, NJ, managed by Chip Bergeron, is located about 40 ft above the condenser and associated balance-of-plant equipment and runs primarily baseload. The collector air intake filters for this unit were not safely accessible without locking out the exciter, regardless of whether the unit was online or offline.
Reason: The busbar location where the power cables to the brush rigging are bolted is directly above the intake filters. After a steam-turbine trip, caused by a collector ground fault during a heavy rain event, the collector’s filters were removed and found caked in carbon dust. Plus, the collector was damp from the rain being pulled through the door seals by the vacuum created by the clogged filters.
The week prior, an online ring grind had been performed which normally would not be a contributing factor except that the dust generated during the grind was being exhausted and then partially re-ingested by the collector filters. Plus, the filters, which are impossible to inspect or service with the plant online, also are subjected to an excessive amount of carbon dust from normal brush wear, because of the recirculation effect. These two issues eventually caused the aforementioned unit trip and maintenance headache.
The problem was resolved by redesigning the air intake assembly and bringing the ductwork down to ground and installing new, easily accessible filters that could be inspected and serviced with the unit in operation. Additionally, the exhaust duct was reconfigured to direct any carbon dust well beyond the new air intake.
With the filters relocated, the overall cleanliness level of the collector housing improved dramatically and water ingestion is no longer an issue. The air intake filters have been removed, inspected, and cleaned several times with the unit online, without any issue. Addition of a differential-pressure gage across the filters has further improved maintenance efficiency by giving the site team a real-time view of filter health.
Doubling down on haz-gas analyzer reliability
At Woodbridge Energy Center, major equipment is located almost entirely outdoors. When the site went commercial in 2016, it was one of the first to use a newly designed aspirated hazardous-gas detection system, which relies on instrument air and an aspirator to pull air samples from two different compartments on each gas turbine through dedicated LEL sensors.
These sensors have the capability to shut down the turbine should two in either compartment go into a state-of-alarm—that is, high LEL readings and/or a loss of sample flow through the LEL detector—at any given time.
Because the detection system is exposed to the elements, issues immediately began to arise with sensor stability on days when ambient conditions changed quickly. The issue was exacerbated when high winds were present. Sensor instability caused numerous false alarms which led to unit runbacks and in a few instances, trips.
Staff learned through testing that the factory-installed stainless-steel cover for the LEL sensors was creating a thermal-sensitivity issue within the sensor itself. This was most evident during sudden rain events in the summer where the cover and sensor temperatures could drop by 20 deg F in a matter of seconds. That temperature drop was even greater (perhaps even 100 deg F) when the rain event occurred right after the sensor cover had been exposed to direct sunlight. Winter brought similar issues where the sunlight would warm the sensor but then high winds would cool it off rapidly and repeatedly.
To solve the problem, the steel cover required protection against the elements. Plant staff and the OEM decided the best option was to double up on the sensor covers. The factory-installed steel cover would stay as is, but a new plastic shield would be clamped around it. Plastic conducts heat very poorly and also blocks most, if not all, the rain and wind from reaching the steel cover.
Since installing the secondary covers, performance of the 24 haz-gas detectors during weather events has been flawless. This simple and relatively inexpensive solution dramatically increased plant reliability.
Color-code plant drains to expedite event response
Woodbridge Energy Center is boarded on two sides by wetlands. The site must properly capture and direct water from different sources—such as blowdown tanks, chemical containments, and various drain sumps. In order to perform this task, Woodbridge uses 128 floor drains and five sumps to direct water to two separate locations (cooling-tower reuse or the local sewer authority).
Additionally, storm-water runoff is captured by large drains which empty to storm-water basins. Those basins then drain into a retention pond which discharges to the adjacent wetlands.
At face value, this is a typical configuration for outdoor powerplants; but Woodbridge was faced with a problem when it came down to how to properly react to an accidental release (chemical, oil, etc). Because the floor drains were not labeled, it would not be readily apparent (in the moment) what sump should be shut down to prevent the product from escaping to the larger systems and potentially the environment.
This was a significant concern because the sumps, which can only be shut down manually, could easily be several hundred feet away and/or obscured by a building or piece of equipment. Shutting down all sumps would be impractical and time-consuming. The site team had to develop a way to easily and rapidly identify what sump to shut down should a release event occur.
To expedite the drain/sump identification process, plant personnel developed a color-coded chart, which was broken down by sump. The color codes were then used to create permanent signs for each sump, calling out its name and discharge location. After the signs were in place, an identifying color-matched circle was painted next to each drain. The color corresponds to the sump which the drain goes to. Now, if an accidental release occurs, the plant team will instantly know which sump to secure, saving precious time.
Progress in driving towards cultural excellence
As Woodbridge Energy Center entered its fourth year of commercial operation, management wanted to focus on understanding how to reinforce the best elements of the plant’s culture. The challenge the team faced was how to effectively gather and present information in a way that provided understanding and visibility of individual program elements and how they connect the team as a whole.
In November 2018, plant personnel were surveyed and asked to list their top five personal values and rank them in order of importance to themselves. The survey was anonymous to ensure that employees felt comfortable enough to put down their true top five. Once all of the results were collected, the responses were compiled, combined, and sorted in order to come up with the top five guiding values for the entire team.
For Woodbridge, those top five values were teamwork, accountability, integrity, learning, and respect. Results in hand, the team, at a December 2018 safety meeting, came up with the three areas they felt would most benefit from knowing team values—safety, excellence, and availability. These were promoted on the graphic, which then was posted in various locations around the site (conference room, control room, etc) as a reminder of what the team uses to guide and drive daily decision-making.
During the six safety meetings in 2019, each of the five values was singled out and discussed in detail. The discussions focused on explaining what each value truly means to plant personnel, how it applies to the work at hand, and in most cases, a group activity or survey in order to continue building a positive work climate. Surprisingly, the deep dives on team values proved the most beneficial and rewarding part of the process.
This was particularly evident during the February 2019 meeting where the value of learning was highlighted. Prior to the meeting, plant personnel were given a 70-question Learning Styles Survey and asked to answer each question from 0 (not like me) to 2 (exactly like me). Survey results were averaged across all 22 team members to create a radar plot showing how the team learns best. Using this information, management was able to cater to the remaining values discussions in a way that would be the most beneficial to employees.
The learning-styles assessment allowed development of a teamwork exercise that encompassed the styles of all members. These learning styles were taken into consideration when creating future training programs for the full plant staff. This personalized approach to the values discussions, safety meetings, and training in general continued on through the remainder of 2019. It culminated in a year-end values survey to reflect on how this experience improved staff morale while creating a sense of ownership.
After the values assessment on learning styles, it was possible to plan training for the remainder of the year that best met the needs of the plant staff. This intentional planning led to more successful training and higher concept retention.