Coal, Gas

Reliable Superheater Tube Wall Temperature Control for HRSGs with Duct Firing

Issue 2 and Volume 118.

By Ilya Yarinovsky, Boiler/HRSG Sr. Specialist and Don Koza, PE, Principal Engineer, Bechtel Power Corp.

The recent technology innovation that has allowed recovery of vast amounts of natural gas from shale deposits has resulted in very competitive (low) pricing of natural gas for use in electric power production. This in turn fueled a boom in the construction of combined cycle power plants (CCPP), which currently represent over 41 percent of all installed electric power generation nameplate capacity in the U.S. and almost 40 percent of net summer capacity. They are historically used as peaking plants and frequently operate in cycling mode (off at night and during weekends); however, the plants now tend to be called upon for following the load demand and even for baseload operation. These various modes of operation impose a demand for high reliability of the major components, such as steam and gas turbines and heat recovery steam generators (HRSGs). Frequently, HRSGs are equipped with duct burners to increase overall seasonal flexibility of CCPPs as well as to support high steam demands of cogeneration plants. For combustion, duct burners utilize residual oxygen available in the gas turbine exhaust.

Like every sophisticated component, the duct burners are sensitive to the operating environment. Deviations from the design conditions may cause combustion instability, deterioration of the mechanical components and increased emissions. However, without proper control, the duct burner may impose the most severe damage on the HRSG – flame impingement on downstream heating surfaces causing creep damage of the pressure parts, and occasionally catastrophic failure of the affected tubes.

Unreliable control of the tube wall temperature may either hinder performance of the HRSG, or, as an extreme case, cause overheating, resulting in subsequent tube failure and lengthy and costly outages as a result.

Gas side

Duct burners designed by the major suppliers consist of a number of burner elements equipped with flame stabilizers (sometimes called flameholders) promoting vigorous mixing at the root of the flame. A burner element consists of a hollow alloy pipe, or “runner”, with orifices spaced along the length.

The flameholder, with or without perforations, resembles a “bat wing” and enhances combustion efficiency. A system of baffles designed to ensure a proper “approach velocity” of the exhaust gases is typically located in spaces between burner runners.

Duct burners are typically located in the inlet duct for smaller HRSGs, or in a dedicated burner duct for large units used on combined cycle power plants. The requirement for the uniformity of the gas turbine exhaust stream entering the duct burner is very high, therefore the duct burners located in the inlet ducts are equipped with distribution grids.

In order to evenly distribute the gas turbine exhaust flow to the duct burners, the distribution grid introduces a slight pressure drop. A similar approach is used for the burner ducts, where an upstream tube bundle generates a slight pressure drop as it provides even distribution of the incoming gas turbine exhaust flow.

The gas turbine exhaust flow characteristics, velocity and temperature distribution are strongly influenced by the design specifics of the gas turbine exit plenum, as well as the configuration of the HRSG inlet duct. These parameters have a strong influence on overall performance and reliability of the duct burner.

Each burner original equipment manufacturer (OEM) provides strict requirements on the flow velocity distribution entering the duct burner. It is not uncommon to see the duct burner OEM’s requirement to have velocity peaks deviating from the mean exhaust gas velocity by no more than ±15 percent over at least 85 percent of the burner inlet plane cross section. Deviation from this requirement would cause longer flames and, as a result, a potential for localized hot spots on the downstream heating surface and even on the sidewalls. HRSG suppliers cannot always comply with this design criterion for the flow distribution devices. To satisfy this condition, the distribution grids (or tube bundles) should be designed for pressure drops well in excess of the current industry standard of 0.5 – 1.0 inch of water column, thus detrimentally affecting combustion turbine performance.

Typically, the HRSG manufacturers provide pad welded thermocouples installed directly on the tubes of the heat transfer section immediately downstream of the duct burners for continuous monitoring of the tube metal temperature and detecting temperature spikes. Although this approach appears fairly reliable at first glance, because of economic and space management considerations these thermocouples are typically installed on just 10 percent of the tubes, leaving the remaining 90 percent without any temperature indication. In the extreme case, such as 100 percent duct burner firing at partial combustion turbine load, when the steam maldistribution will, in all likelihood, be coupled with a temperature maldistribution on the gas side, more reliable monitoring will be required.

In addition to the pad welded thermocouples located on the tubes of the heat transfer section, HRSGs are frequently equipped with a grid of thermocouples to measure the flue gas temperatures entering the heat transfer surface facing the burner. These thermocouples should be strategically located in the duct, capable of measuring temperature peaks and sufficiently shielded from the heat radiated from the flame in order to provide an accurate flue gas temperature reading. Power plant controls should be adequately designed to ensure duct burner run back or shutdown when measured temperatures exceed the maximum allowable limits for the heat transfer tube material.

The thermocouple arrangement widely used in the industry typically consists of three or four thermocouples penetrating each side of the burner duct at different elevations. Each individual thermocouple is inserted into a protection tube made of austenitic stainless steel to resist elevated gas temperatures. Depending upon the length of the protection tube, it may be braced for stiffening against vibrations caused by vortex shedding.

It is very important to obtain the most reliable indication of the temperature peaks downstream of the duct burner. The location of the flue gas side thermocouples should be selected based on the results of either a physical scale model study or computational fluid dynamic (CFD) simulation which is capable of predicting the temperature distribution.

Based on the results of the temperature distribution studies, one of the following thermocouple arrangements should be adopted.

Arrangement 1 – for double- or triple-wide HRSGs (Fig. 1): This arrangement uses a total of ten thermocouple/protection tube assemblies (five on each side) penetrating the side walls of the ductwork immediately upstream of the heat transfer section. Each protection tube should be equipped with a radiation shield. The length of the assembly should be adequate for placing the tip of the thermocouple in the highest temperature zone at each elevation as predicted by the flow model study.

fig. 1

Additionally, at least six thermocouples should penetrate the roof (or floor, depending upon the HRSG configuration) panel of the burner duct to detect the exhaust gas temperature at the exit elevation of the heat transfer module downstream of the burner, as this point corresponds to the highest steam temperature and the monitoring of exhaust gas temperature at this elevation is critical for reliable operation of the HRSG. It should be noted that the control system should operate based on the setpoints established by each individual HRSG OEM.

Arrangement 2 – for double- or triple-wide HRSGs (Fig. 2): A total of five perforated stainless steel pipes should span across the full width of the HRSG at elevations selected by the HRSG OEM based on the results of the flow model study. At least two thermocouples should be inserted at different depths at each side of each individual perforated pipe. Thermocouples detecting gas temperatures corresponding to the highest steam temperatures as described above will also be required.

fig. 2

It should be noted that the accurate flue gas side temperature control alone cannot guarantee safe and reliable operations of the superheater tubes. It is very important to ensure that the superheater module is designed such that each tube receives adequate steam cooling flow on the inside. Proper steam flow distribution inside of the heat transfer section coupled with relatively uniform flue gas temperatures on the outside will ensure that the tube wall temperature will remain below its design limit for the duration of the HRSG’s life.

The following section provides a methodology for verification if proper internal flow distribution would be established in the HRSG modules.

Steam side distribution in the superheaters

The exhaust gases have a temperature range between approximately 1200°F to 1700°F entering the first heat transfer sections of the HRSG. Depending upon the design of each individual HRSG, the first section facing the duct burner may be either superheater or reheater, or, in some cases, evaporator. The majority of the HRSGs are designed such that the first heat transfer section facing the duct burner is the superheater.

A typical superheater consists of two horizontal headers connected with finned heat transfer tubes (Fig.3). Steam enters a supply header, travels in the horizontal direction as shown, and then enters vertical tubes. In an ideal situation each tube will receive an equal amount of steam flow. However, the hydraulic distribution strongly depends on the supply header module geometry and the arrangement of the supply/discharge piping vs. heat transfer tubes. Steam flow through the individual heat transfer tube is determined by the static pressure differential in the supply and discharge headers corresponding to each physical location along the length of the header relative to the supply and discharge connections in the header. This pressure differential may be different for various steam supply, distribution and discharge schemes.

fig. 3

It should be noted that the boiler industry generally uses two types of such schemes: i) a radial supply/discharge scheme where the supply connection is at the center of the header section and flow radiates in both directions and the flow is likewise gathered and drained through a central connection in the drain header; and ii) a scheme with a side (header end) supply and discharge (Fig. 4). Obviously, different combinations of these arrangements are also widely employed by the industry.

fig. 4

The following example demonstrates a typical steam flow distribution concept for a module with Z-type discharge and supply arrangement, as shown in Fig.5 below.

fig. 5

The Bernoulli equation for the supply header, considering header friction losses, can be written as:

eq. 1

Where

P1 and PL = static pressures in the cross sections corresponding to the left and right end plates of the supply header respectively

V1 and VL = steam velocities in the subject cross sections

ρSH = Superheated steam density in the supply header

dPSH_Loss = friction losses in the supply header

Obviously the velocity of steam traveling along the header will be falling as the steam is supplied to the tubes, and, at the plane corresponding to the right end of the header, VL = 0

Therefore, the Bernoulli equation can be re-written as follows:

eq. 2

This means that the pressure at the far end of the supply header will be greater than the pressure at the inlet by the value of dynamic head less the header friction losses.

The Bernoulli equation for the discharge header is represented below:

eq. 3

Where

P0 and P2 – static pressures in the cross sections corresponding to the left and right end plates of the distribution header respectively

V0 and V2 – steam velocities in the subject cross sections

ρDH = Superheated steam density in the discharge header

dPDH_Loss = friction losses in the discharge header

Obviously, V0 = 0 and the equation (3) can be re-written in the following form:

eq. 4

The equation (4) shows that the left end of the discharge header will have the highest static pressure.

When we compare equations (2) and (4) it is obvious that due to the static pressure differential between the discharge and supply headers, the tubes at the left side of the module will receive less cooling steam flow. This will increase the module’s susceptibility to tube overheating and potential failure, especially if the gas temperature facing the superheater reaches its peak at the location corresponding to the tubes receiving the least flow.

The above analysis is presented for the Z-type arrangement due to the relative simplicity in demonstrating the concepts governing the steam flow distribution in the HRSG superheater/reheater modules. However, this concept can be adopted for different types of both radial and side steam supply and discharge arrangements, and, as a matter of fact, the HRSG suppliers are using this concept tailored to their individual state-of-the-art designs.

It should be noted that the steam distribution through the SH module will be substantially aggravated if the tubes’ pressure drop is low enough to be comparable with the pressure drop through the headers. The designer should carefully evaluate the flow distribution, especially for the single pass superheater and reheater modules. It is obvious that in order to achieve an acceptable steam distribution, the pressure drop within the tubes should be significantly higher than one in the header. Typically, the negative effects of the flow distribution are negligible for the two-pass superheater modules, as the tube pressure drop is substantially higher than the header loss, even at low loads.

Conclusion

A single temperature measurement of the steam exiting the heat transfer section, precise measurement of the gas side temperature peaks and a theoretical correlation between the gas, steam and tube wall temperatures provided by OEM, will ensure a reliable control of the superheater temperature regime.

The organization of superheater tube wall temperature control can be reliably addressed with the following combined measurement/analytical approach:

i) The HRSG should be equipped with strategically located gas side thermocouples located immediately upstream of the superheater with the highest outlet steam temperature, which faces the duct burner. The location of the gas side thermocouples should be established based on the results of the Flow Model Test and/or CFD analysis;

ii) The Heat Transfer Section facing the duct burner should be designed with a uniform steam flow distribution, which can be achieved either by the pressure drop, or by maintaining a certain cross sectional area ratio between the collection header and the attached tubes.

iii) The OEM should develop a correlation algorithm between gas / steam / tube wall temperatures for implementation in the DCS

This approach has a good potential to ensure a cost effective and reliable approach to tube wall temperature control across the range of operational loads.

References

1. “Standard Methods Of Hydraulic Design For Power Boilers”, Taylor & Francis; 1 edition (January 1, 1988)

2. “Kotel’nye Agregaty” (“Steam Boilers”), by M.A. Styrikovich (1959)

More Power Engineering Issue Articles
Power Engineerng Issue Archives
View Power Generation Articles on PennEnergy.com