By Robert L. Hadley, Progress Energy - Carolinas & Larry Trom, Alfa Laval, Inc.
Unit 1 of Progress Energy’s Lee Plant, in Goldsboro, NC, was installed in 1952 with a Westinghouse RT646 turbine rated at 84 MW. As was standard for that time, the lube oil coolers were shell-and-tube type heat exchangers, physically located within the turbine sump. The two, 100 percent capacity units were originally specified to operate with an inlet water temperature of 900 F, coming off the cooling tower. Each consisted of 976 5/8” diameter admiralty tubes, five feet in length with a surface area of 850 sq.ft.
The original units were kept in operation until 1994, when they were retubed as a result of lost surface area from plugged tubes and frequent maintenance. However, within two years each unit again began to suffer from leaks. Repairs were being made at least once a year on the system over the next eight years. In 2003, a serious leak developed, for which the unit had to be retired and an outside contractor brought in to make repairs.
 Photo 1. View of layout and water-side taps from the existing piping. Photo courtesy Alfa Laval. |
At the same time, the unit’s concentrated cooling water (CCW) system was performing at far less than original capacity as a result of cooling tower deterioration over the years. Thus 90 F was an optimistic temperature during the hottest times of the year and lube oil temperatures were missing design conditions of 115 F by a wide margin.
Progress Energy had alternatives. They could continue to spend money repairing the coolers every year, but that wouldn’t solve the temperature issue in the summer. They could rebuild/replace the cooling tower to recover the temperatures. That would be a significant expense and they would still need to repair the shell-and-tube coolers every year. Or these could decide to redesign the system. Progress decided to redesign the system.
Several years earlier, Progress had built a full-flow oil flushing skid, which was used during outages to thermally shock piping for de-scaling. A plate heat exchanger was used on the skid to heat and cool the shock fluid. The compact size of the plate heat exchanger allowed for a much smaller skid. As the plate exchanger had performed without problems, Progress decided to consider that technology as a replacement option to the shell-and-tube coolers.
Plate Heat Exchangers
Plate heat exchangers (PHEs) have gained acceptance in all industries today, with a global annual sales volume of over $1 billion. They are chosen because one or more of their advantages fit a given application or solve a particular problem for a customer. These advantages include lower capital cost, lower installation cost, greater heat rejection, ability to use warmer circulation water, small space requirements and the ability to be disassembled, moved into location and then reassembled.
A gasketed PHE consists of a series of thin, corrugated alloy plates separated by gaskets and bolted together in a carbon steel frame. The plates and gaskets are arranged so that hot and cold media flow alternately in each channel. Figure 1 illustrates the operating principle of a gasketed plate heat exchanger. The PHE achieves the highest “U” values of all the compact heat exchangers. It can be disassembled for cleaning and has the flexibility for plates to be added or subtracted should application requirements change. It can also be disassembled for moving into the installation site, an important feature when replacing large shell-and-tube exchangers such as those in CCW service located within existing structures.
 Figure 1. Exploded view of a gasketed plate heat exchange (PHE) showing flow paths for both hot and cold liquids. Illustration courtesy Alfa Laval. |
The wetted parts of a plate exchanger are normally supplied as alloy, with 300 series stainless at a minimum. Because of the units’ high efficiency, however, the cost of a stainless plate exchanger is usually much lower than that of a carbon steel shell and tube (S&T), and thus lower than copper or higher alloy tubular exchangers.
Plate exchangers can also achieve a temperature cross and very close temperature approaches. Either one is difficult with an S&T as the size and complexity increase tremendously.
The Decision
Like most users, Progress Energy’s first thought was simply to retube the existing exchangers. This would have required removing the bundles, shipping them to a facility and reinstalling them. The estimated cost for retubing was approximately $89,000 for both bundles. Site removal, shipping and installation would have been additional. This would have solved the recurring leaks, but based on the most recent (1994) retube, problems could again arise in two years. This also would not have solved the high lube oil temperatures resulting from the aging CCW system.
Based on the experience with plate exchangers on the flushing skid and elsewhere in Progress’ system, Alfa Laval was asked to estimate the cost for new plate exchangers to replace the shell-and-tubes.
Two estimates were made. The first was to match the original thermal rating of the existing exchangers. The second was to increase the inlet CCW temperature to 105 F (which is a possible summertime temperature with the current condition of the cooling tower) and reduce the final oil temp by 5 F to 110 F.
The results of the design study are presented in Table 1.
 |
Several facts are evident from the data.
- Lower oil return temperature - The new system case in a PHE was able to cool the oil to 110 F even though the incoming CTW temperature was increased by 15 F.
- Less water usage - In both cases, the plate exchanger uses less water than the S&T. Because of its ability to use a temperature cross and achieve closer approaches, Design 2 uses half the flow of water, leaving more for other uses. At this site, because of high CTW temperatures, additional water for other equipment will better cool those services
- Higher efficiency - The efficiency (comparing U-values) of the base PHE design is nearly twice that of the shell-and-tube heat exchanger. This is true even though the temperature program suits an S&T exchanger and is far from ideal for a PHE.(S&T exchangers perform at their best when there is no temperature cross and where there is a large temperature approach, as in the original design.) Design 2 is nearly three times as efficient.
Before a decision could be made, other factors had to be determined. The existing exchangers were located within the sump. If plate exchangers were to be used, a new location had to be determined, tie-ins to both the lube oil system and the water supply had to be planned and costs had to be compared.
The site was physically reviewed by Progress Energy’s engineering, operations and maintenance groups and a suitable location next to the sump was found. Although head clearance was relatively low, the proposed exchangers’ profiles allowed not only for setting in the desired location, but delivery through a standard 36-inch building doorframe.
System Design
Water supply to the existing exchangers was from individual supply and return lines through the floor. It was decided that one set of lines would be left in place; the other would supply the new exchangers. Oil supply would come from pump discharge and return to the bearing header with both connections made through the sump wall. Connections on this model plate exchanger were standard at 6-inch diameter, studded ports. (Studded connections are used on plate exchangers as they allow the frame to be used as a pipe support, with much higher loads possible than with extended nozzles.) Thus, 6-inch carbon steel pipe runs would be used. (In retrospect, savings could have been achieved on the pipe and valves had 4-inches been used with a 6 x 4 reducer at the exchanger. Pipe runs were short and would experience little head loss with that size.)
As in the original sump, two 100 percent exchangers would be used for redundancy. It was decided to use a dual transfer valve on the oil lines to be able to switch oil flow from one exchanger to the other with a single manual level controlling both supply and return. This would eliminate any possibility to interrupt oil flow to the turbine by oversight during changeover. Water lines would be connected in a more standard manner using block valves. In the final installation, block valves were added to the oil lines as well, as a backup to the transfer valve during cleaning of a unit while the other is online.
A standard plate exchanger has much tighter passages than a shell-and-tube, and it’s advisable to use a strainer if debris is a concern. Lake and river water is routinely used in plate exchangers with a proper strainer. Cooling tower applications may require strainers as well if large amounts of bugs, leaves and other debris are present. At the Lee Plant, there was a concern that years of corrosion and scale buildup might loosen with the disturbance, become mobile and cause plugging. A cone strainer was included in the final installation.
 Photo 3. Installation view of sump connections (left) and transfer and block valves on oil-side (center). Photo courtesy Alfa Laval. |
Temperature and pressure gauges were installed in front of and behind each unit so proper measurements could be made. Periodic review of pressure drop and temperature will indicate any drop in performance. This would be an early indicator that a cleaning of the affected unit should be scheduled. A reduction in thermal performance would indicate scale or biologic growth on the plates; an increase in pressure drop would indicate plugging due to debris.
The plate exchanger retrofit costs compared favorably to the shell-and-tube option. The total cost for the plate exchangers, the transfer valve and the complete installation including setting and piping modifications came to approximately $80,000. This system solves several problems: 1) it can reduce oil temperature 5 F below the original design condition; 2) it can accomplish the duty with 15 F higher water temperature and therefore not require cooling tower work; 3) additional water was made available for use by other components. Retubing the existing units would have cost close to $100,000 when all costs were included and would not have solved the three items above.
Performance
The system was put on line in the winter of 2004-2005. The desired oil temperature of 110 F was easily achieved with the water control valve only cracked open. Parameters at the time of this writing are presented in Table 2.
 |
There has been no maintenance performed on the equipment during startup or operation to date, nor is any expected by Alfa Laval. The supplier expects the gaskets to last 10 years or more at the lube oil temperatures indicated. Plates are expected to last 30 years at the current water conditions. Units need be opened only if performance, as indicated by temperature or pressure, indicates a problem. With stainless steel wetted parts, inspection for corrosion should be an infrequent procedure.
Authors: Bob Hadley is Principal Engineer, East Region, for Progress Energy. Larry Trom is Market Development Manager, for Alfa Laval Inc.