JEA electric utility saves $206,466 annually through improved monitoring of condenser performance

Sentry System helps optimize plant efficiency by precisely detecting air-in leakage in turbine exhaust systems.

October 24, 2001 — Just as the days of juicy “cost-plus” defense contracts have gone the way of pocket protectors, so too have the days of running a utility like a government bureaucracy.

A guaranteed flow of funds is drying up and electric utilities are now forced to run a tight ship with an ever-vigilant eye on the bottom line. Every dollar counts, and every engineering efficiency must be explored to stay competitive.

One U.S. power supplier managed to realize six-figures of annualized savings through accurate monitoring and control of their turbine condenser exhaust system. Thanks to the incorporation of an accurate gauge to determine air-in leakage, Jacksonville Electric Authority (JEA) of Jacksonville, Florida also profited from intangible savings such as decreased maintenance and improved boiler reliability. JEA’s experience sets a prime example for other utilities that wish to improve business operations by quantifying, analyzing, and then optimizing their power-generation processes.

JEA is the eighth largest municipal utility in the United States. It currently provides electricity to more than 355,000 customers in Jacksonville and parts of three adjacent counties in Florida. With a net generating capability of 2,361 megawatts, JEA wholly owns and operates three generating plants and all transmission and distribution facilities.

JEA’s Northside Generating Station operates two gas/oil-fired steam units, the largest of which, Unit #3, is a 520 MW unit with a Westinghouse steam turbine and a De Laval steam condenser. The air is pumped out of the condenser by a Nash vacuum pump that discharges to the atmosphere.

Knowing that this unit had operated continuously since installation in 1979, and prompted by upper management’s drive to improve plant efficiency, Fred Maner, one of JEA’s senior engineers, took up the challenge to focus on the condenser system for extracting further gains in electricity output.

“My company was looking for ways to improve our thermal performance in the plant, and I immediately set my sites on the condenser,” recalls Maner. “It represents the largest heat-loss source in the steam cycle, so it must operate efficiently. Reducing the air-in leakage on the steam side is particularly critical to improving overall condenser performance. If the condenser is not running efficiently, typically that means that you’re running at a higher back pressure which results in decreased turbine performance.”

Quantitative analysis measured in dollars and cents

Condenser air in-leakage is notorious for sapping plant capacity and efficiency. An increase of one inch of mercury in condenser backpressure can conceivably limit unit capacity by as much as 2%. The heat-rate penalty can be just as severe, approximating 16 Btu/kW-Hr for every 0.1 in. Hg Abs. increase in turbine backpressure. Air in-leakage also adversely affects a generating unit’s water chemistry balance by increasing dissolved oxygen, ammonia and carbon dioxide levels in the condensate�resulting in heat exchanger and boiler corrosion problems. Over time these losses combine to grossly inflate operating and maintenance expenses.

In light of the potential gains to be had by reducing air in-leakage, Maner and his team began a gross analysis of the Unit 3 condenser air removal system in May of 2000. Initial examinations revealed that existing process-control metrics were inadequate.

Historically, a pressure gauge is used to decide when turbine back pressure is too high. In conjunction with the pressure gauge, a rotameter is used periodically to check air in-leakage, a primary contributor to excess turbine back pressure. However, due to daily plant fluctuations, changes in back pressure are only recognized in increments of 0.3 inch Hg Abs or more. Yet, Maner calculated that an increase in turbine back pressure of just 0.1 inch Hg Abs. added up to $511.00 per day in lost energy to JEA. Given this pernicious financial drain, the need for a more sensitive and accurate monitor took on added urgency.

During the Fall of 2000, Maner was approached by Joseph Harpster, president of Intek, Inc., to try out a new system for measuring air-in leakage.

“What prompted us to try Intek’s new Sentry system, is that Joe Harpster and I had already done quite of bit of work together,” says Maner. “I originally bought one of their RheoVac®, single-probe monitors and installed it the on the Northside #1 generator. So when Joe developed this new system, he decided that it would be a good idea to allow me to put it in operation. Since the Unit 3 generator was currently under close scrutiny, we jumped at the chance to install the Sentry system there.”

Based in Westerville, Ohio, INTEK, Inc. has been manufacturing its Rheotherm® line of flow instruments since 1978. These instruments are used worldwide in a variety of industries for precise liquid and gas flow monitoring. Intek’s RheoVac Sentry system uses this mass flow meter and three other gas property sensors to perform its unique measurement. This system enables users of large vacuum-based water extraction systems to maximize system efficiency by providing multiple, strategically placed probes to monitor vacuum performance.

These very sophisticated RheoVac sensors are built into a 316 stainless steel probe which installs through a threaded, 11/2″ hot tap assembly into 3″ to 16″ vacuum exhaust lines. There are no moving parts.

“One of the drawbacks to using a single sensor, such as a rotameter, is that you only have one origin of data,” points out Maner. “But with a multipoint system like the Intek Sentry, you can do summation calculations around the various probes and it will give you a quick double-check.”

Installation of an accurate monitor produces valuable revelations

JEA installed Intek’s Sentry system in January of 2001.

“On Unit 3 we have two dual pass, divided waterbox, condensers side by side, however, they’re tied together, so it’s single pressure, not multi-pressure like some of the newer units,” says Maner. “On each condenser neck there are two air-removal headers that come out of each individual tube bundle. Then they’re all tied together into one larger header. Each one of these tube bundle air-removal headers has a RheoVac probe in it. Where those two are tied together, we have another validation probe. So we have 6 probes along the two compartment condenser fronts, and then a seventh probe in a final header to measure the summation or total flow out of the condenser.”

To ensure that the Intek Sentry system would accurately provide the necessary data, a gauge study was initiated. The results proved the probes to be sufficiently accurate, with a part-to-part variation of just 94.04% within 6 distinct categories.

For determining the efficiency of the condenser air removal system, Maner and his team selected air in-leak flow as the primary metric, and turbine backpressure as the secondary metric.

By measuring the air in-leak flow as measured upstream of the Nash vacuum pump, a high-load, air in-leakage flow baseline was established at 40 scfm. During low load operation, air in-leak flows above 100 scfm were common. The Sentry system also revealed a turbine backpressure of 2.41 Hg Abs.

These figures indicated that the Northside Unit 3 condenser was out of specification compliance by 25 scfm. Maner then extrapolated that the excess air in-leakage at full load was costing JEA approximately $206,466.00 per year in extra fuel costs and much greater lost revenue in unrealized power generation.

Effecting a money-saving solution

Armed with conclusive evidence of excessive leakage, Maner ‘s next step involved determining the scope of necessary work to reduce the number of air in-leak sources.

Maner elected to use Helium leak detection equipment, as it is an industry standard and a highly effective method for resolving air in-leakage problems by locating the source.

Culprit sources turned out to be water traps from the boiler feed pumps, feedwater booster pumps, the steam packing exhauster, the LP turbine to condenser expansion joint, the LP turbine rupture disks and the LP turbine steam seals.

A final work-scope was generated based on the statistical data analysis and the results of the Helium leak detection testing. Earmarked improvements included: replacement of the water traps with a new receiver tank design; repairing the LP turbine rupture disks; repair of the turbine-to-condenser expansion joint; and major rebuilding of the LP turbine steam seals and housings.

With a large return on investment in mind, JEA management wisely released the necessary capital to remedy most of these shortcomings. By March of 2001, much of the necessary repairs were made.

“The water traps on the boiler feed pumps, feedwater booster pumps, and the steam packing exhauster were replaced with one receiver tank,” says Maner. “This eliminated five water traps and provided a superior liquid seal between the condenser and the atmosphere. Rebuilding the steam seals and clearances was delayed only because that work will require a major outage to complete.”

“By the end of March 2001, the air in-leakage dropped below the upper specification limit,” notes Maner. “Our new high load baseline is approximately 13 scfm, and during low load operation the air in-leakage flow typically remains below 25 scfm. The overall measured turbine back pressure reached 2.13 Hg Abs., a drop of 0.3 Hg.”

Comparing data gathered via the Intek system from April 2001�after the corrective measure were undertaken�against data from April 2000, Manor calculated a 67.5% and 13.4% reduction in the air in-leak flow and turbine back pressure respectively.

Important to JEA’s finances, these engineering improvements accounted for $206,466 in annual savings due to reduced turbine backpressure. In addition, intangible savings stemmed from improved boiler cleanliness and reliability. Maner estimated that boiler chemical cleanings range in cost from $300,000 to $500,000, depending on unit-specific requirements. Since a reduction in air in-leakage means less corrosive product is carried over to the boiler, then savings result because the mean time between cleanings is improved. Even more impressive, big-ticket savings could be realized if just one furnace-tube failure event was eliminated due to a reduction in air in-leakage during a peak demand period. Maner figured that up to $2,246,400.00 in expenses could be avoided by preventing one single such event.

Implementation of control feedback to maintain savings

Bolstered by immediate improvements to operations because of Intek’s Sentry system, JEA’s engineering staff insisted on incorporating this monitoring system into the condenser control system loop. In this manner, progress could be realized under all demand conditions.

“The production of electricity is a true ‘pull system,'” observes Maner. “Our process must be flexible in order to meet customers’ continuously changing specification limits. In the power generation business, we operate, maintain, control and manage variation.”

JEA’s current plan incorporates three control measures: a statistical process control chart, a daily water chemistry report and a plant information process book. Data for the air in-leak process control chart is gathered during each month at both high and low load conditions. The daily water report indicates the generating unit’s overall condition. The process book is a time series data trending tool, used to identify and troubleshoot system problems. The book is also a valuable tool for process trend analysis.

“We expect to further reduce the air in-leakage and meet or exceed our original target values when the LP turbine steam seal and turbine steam seal supply system is completely repaired,” says Maner. “With the continuous monitoring of the condenser system, we can immediately quantify the further gains in turbine efficiency at that time.”

The success of Maner’s air in-leakage reduction project spawned the quest for ferreting out additional areas for improving plant efficiency.

“During the measure and analysis phases of this project it became apparent to me that there is a significant problem concerning the turbine steam seal system,” says Maner. “Numerous high-energy steam sources from this process are dumping to the condenser. These high energy over-board flows are directly impacting turbine and condenser performance, so I requested approval to next pursue a solution to this problem.”

Cognizant of the financial gains realized by Maner’s first project, JEA’s management quickly granted approval to initiate the second phase of optimization.

“The goal of the original project was to reduce the excessive air in-leakage and improve turbine back pressure on Unit #3 by July 2001,” says Maner. “The project successfully accomplished that goal three months ahead of schedule and the process is now complete. I’m certain we will have equal success on our next project.”

Maner credits much of the success at JEA on having valuable data to work with.

“Just having the right measurement capability is really the key,” says Maner. “The Sentry gives you quite a number of trouble-shooting improvements. Especially when you have a condenser with a lot of potential sources for leaks, it helps to know where and how much of a leak is occurring.”

For more information about the RheoVac Sentry system, contact Intek, Inc. at e-mail; or visit the web site:

Previous articleValero Energy Corporation reports lower third quarter; income still higher for year
Next articleABB reports improved operating performance

No posts to display