face=+Italic; By design, combined HP-IP turbines have a small amount of internal leakage from the high-pressure turbine to the intermediate-pressure turbine. As turbines age, the leakage increases considerably and becomes excessive, creating a heat rate penalty and possibly a reliability problem. Last month we explored the symptoms and causes of steam leakage within GE steam turbines and how to correct the problem. In Part II, we examine the same issues for Westinghouse and Allis-Chalmers turbines from both theoretical and practical angles.face=-Italic;For Westinghouse (Figure 1) and Allis-Chalmers combined high-pressure/intermediate-pressure (HP-IP) turbines, there are more design (and potential nondesign) leakages to the IP turbine than in General Electric machines. That's because the former units require balancing the thrust loading of the reaction blading. Leakage can enter the IP turbine inlet, IP turbine exhaust, low-pressure (LP) crossover pipes, and occur between split IP1/IP2 turbines. Its result can be an apparent increase or decrease of IP turbine efficiency, depending on the source.

Southern Company Generation (SCG) has found that trending of the apparent IP efficiency using the Booth-Kautzmann temperature split tests (discussed last issue in Part I) is very useful for resolving internal leakage problems on our Westinghouse and Allis-Chalmers turbines. It should be noted that the true IP blade path efficiency of these units is difficult to measure.

For later (BB44) Westinghouse turbines, the sources of design leakage (Figure 2) include:

Main steam inlet bell seals (the leakage bypasses the HP-IP turbine; more recent units indicate zero leakage on the design heat balance).

HP-IP-LP dummy balance seals (Figure 3).

HP exhaust piston seal rings.

IP inlet (hot reheat) piston seal rings.

Depending on the unit design, HP dummy seal leakage can affect HP turbine efficiency as it flows through the equilibrium pipes to the HP exhaust. If the IP dummy ring is severely distorted, it is possible for the IP dummy leakage to be a mix of HP dummy leakage and HP exhaust steam flowing in the reverse direction through the equilibrium pipes.

Among the sources of nondesign leakage are:

The HP inner shell horizontal joint (distorted shell or loose bolting).

Broken internal equilibrium balancing pipes.

A broken first-stage drain or pressure-sensing line between the inner and outer shell.

Piston seal rings on the IP turbine mid-point extraction pipe.

A missing mid-span balancing port plug in the inner shell.

Cracks in the main steam inlet pipes at the trepan radius.

face=+Bold; Relating efficiency to leakageface=-Bold;

For the BB44 design, two values of apparent IP efficiency should be measured. The IP efficiency measured from the hot reheat to the IP exhaust at the blading only takes into account the leakage effect of the IP dummy. The IP efficiency measured from the hot reheat to the LP crossover includes the effects of all HP to IP leakages. The temperature split tests are also used to estimate the IP dummy leakage and the total HP to IP leakage.

Although the test wells provided by the turbine manufacturer at the IP blading exhaust on the upper and lower (difficult-to-access) shells are in stratified locations, they are still useful for trending. Because the deaerator and boiler feedpump usually extract at these points, the temperature of the extraction line also should be measured. On one Southern Company unit that had been retrofitted with a new design IP dummy ring, we also installed two new test wells in the upper outer cylinder just upstream of the hot reheat inlet pipes. The results of testing using these wells still indicated some stratification of IP exhaust temperature.

A broken or cracked bell seal on a BB44 turbine can be diagnosed by trending the difference between the deaerator/boiler feedpump turbine extraction temperature and the LP crossover pipe temperature. If the former is several degrees lower than the latter and the two crossover pipes are at a different temperature, a bell seal or inlet pipe could be cracked.

For units with electrohydraulically controlled governor valves that are individually operated, the manufacturer's bell seal diagnostic test is useful to perform. Each of the eight governor valves is closed (and re-opened) and the above temperatures are trended. (Exercise caution when doing so on the normal initial opening governor valve, including temporarily disabling the unit trip function.) A change in the temperature differences will be observed on a leaking bell seal when the corresponding governor valve is closed; this reduces but does not eliminate the leakage, which is reverse flow from the first stage, rather than main steam. If the testing indicates that an upper bell seal is the problem, it can be fixed relatively quickly because only the upper half of the outer cylinder must be removed. But if the testing indicates a problem with a lower bell seal, fixing it will not be easy because it will require removing all HP-IP turbine components.