Picture This: Efficient Desuperheating Revealed

Energy producers such as district heating, pulp and paper, petrochemical, and desalination plants are constantly striving to boost efficiency and reduce emissions. In part, this can be achieved by improving their steam conditioning equipment.

Factors that can impede producers from reaching these goals include control problems in steam systems close to saturation, the behavior of temperature gauges depending on installation, and a lack of knowledge about droplet behavior, fluid mechanics, and thermodynamics.

Steam producing systems are challenging because they are non-linear and thus difficult for controllers to handle. Transport delay varies with the steam flow, and the behavior of gauges and transducers differs with the state of the steam. Today, the technique is pushed as far as safety will permit. To further improve upon this process, the whole system must be studied in closer detail.

Field studies should be performed to validate simplified, dynamic flow and heat transfer models to examine evaporation time and the behavior of existing and modified installations.

CCI Sweden Begins Field Testing

With its in-depth experience designing and manufacturing equipment for the conditioning and control of steam, CCI Valve Technology AB, located in Säffle, Sweden, is an ideal choice for conducting such field studies. Thus in fall 2003, CCI Sweden began field test evaluations of desuperheater applications by means of thermographic measurements. The tests examined steam conditioning valves and desuperheaters at four different Swedish pulp and paper mills. The thermographic method itself was also evaluated by surface and temperature sensors inside the pipe.

These field tests were intended to evaluate the evaporation process for water injected into superheated steam. Using a thermographic camera to take images of the steam pipe after the water injection point, CCI was able to show how the injected water evaporated and if there was any free water on the pipe wall. The studies also compared the temperature distribution of a simple probe-type desuperheater to a ring-type desuperheater with a liner, such as CCI’s DA-M desuperheater.

CCI’s DA-M Desuperheater

The DA-M is used in desuperheater applications in which large spraywater flows are required for cooling the steam. It is part of the steam line and has several water-atomizing OP nozzles attached to it and connected to a common spraywater pipe. A liner can be installed in the DA-M to improve the system turndown or to protect the steam line.

In the DA-M, the cooling water enters the OP nozzle through admission holes, and the water is rotated around the nozzle plug thanks to the special arrangement of the holes. The angle of the nozzle seat is slightly different from the control plug so that the water velocity will increase during its travel through the nozzle and reach its maximum as the water leaves the nozzle. These two design features – rotation and high velocity – guarantee a fine atomization, which provides fast evaporation of the cooling water.

The basic DA-M design is shown in Figure 1, and Figure 2 shows the actual DA-M installation with the insulation removed. Figure 3 shows the bare pipe exposed slightly downstream from the DA-M water injection point, with the thermographic camera visible in the right-hand side of the photograph.

Figure 1: CCI's ring-type DA-M desuperheater is designed to ensure even distribution and fast evaporation

Figure 2: CCI's DA-M installation with the insulation removed

The Evaluation – CCI’s DA-M Desuperheater

The colors in the thermographic images are chosen automatically by the camera from the temperature span in the image; therefore, photographs of the same pipe can show slightly different colors depending on various background temperatures. The temperature scale should be used to interpret each image individually.

Figure 3: The steam pipe just downstream of the DA-M desuperheater with the insulation removed

In the thermographic image in Figure 4, the steam pipe is clearly visible in red/yellow. To the left and right and under the pipe, an insulation layer is visible. The background shows some equipment being warmed by heat radiating from the pipe. The flow is from right to left and the liner is visible on the right side of the photo.

Figure 4: Thermographic image of steam pipe 1.6 to 3.3 feet (0.5 to 1.0 meters) after the DA-M desuperheater: the liner gives the yellow color

Figures 5 and 6 show temperatures farther downstream of the DA-M. At the bottom of the pipe, a lower temperature is discernible. This lower temperature is much higher than the saturation temperature but indicates that a small amount of wet steam falls down to the bottom of the pipe after the liner and evaporates there.

Figure 5: Thermographic image of steam pipe 3.3 to5.2 feet (1.0 to 1.6 meters) after the DA-M desuperheater: the liner gives the yellow color

Figure 6: Thermographic image of steam pipe 5.2 to 7.2 feet (1.6 to 2.2 meters) after the DA-M desuperheater: the liner gives the yellow color

Assuming that all of the injected water is evaporated when the yellow streak disappears, this corresponds to an evaporation time of 51 milliseconds.

The field test of the DA-M application shows that the atomization and evaporation of the injected water work well, creating no condensate in the downstream pipe. The thermographic photographs indicate that water evaporates quickly, but some wet steam is present just beyond the end of the liner at the lower part of the pipe.

After the liner, the thermographic images from the high load condition show streaks with different temperatures, a typical result of the upstream bend located very close to the desuperheater. The images indicate that the liner has eliminated this problem fairly well. The hottest areas have a temperature of 442°F (228°C), and the coldest areas have a temperature of 376°F (191°C).

In addition to the thermographic camera data, extensive surface and insertion thermo couples were used to verify the temperature profiles. In all cases, there was close agreement between the two independent means of measurement. This data was also used to verify the analytical model used to predict primary and secondary atomization and evaporation.

The evaporation time for the main portion of the injected water in all running conditions is excellent at less than 70 milliseconds. A higher flow yields a longer evaporation time due to the upstream pipe bend.

The Evaluation – Competitor’s Probe-Style Desuperheater

The competitor’s probe-style desuperheater has a variable geometry nozzle design and is used for temperature control of 145 psig (10 bar) steam to the paper mill, paper machines, recovery boiler, and sulfate boilers. The steam comes from a turbine outlet and/or two steam conditioning valves.

The plant has a constant problem with water in its steam pipe and is forced to replace the nozzles in the desuperheater every year. The condition of the steam is crucial for the function and efficiency of both the sulfate boilers and the drying function in the paper machines; a temperature that is too low will result in low efficiency in a sulfate boiler, while a temperature that is too high may cause sticking problems. If damp steam hits the temperature sensors, the process can no longer be properly controlled.

Figures 7 and 8 show the installation of the competitor's probe-style desuperheater. The point of injection is located in a vertical pipe, flow direction up. Downstream there is a 90º bend to horizontal at 13 feet (4 meters) and a 30º bend to horizontal at 28 feet (8.5 meters). The existing temperature sensors are located at 145 feet (45 meters), after two ventilation valves and two other branches.

Figure 7: The steam pipe downstream of a competitor's desuperheater, flow direction up, with the desuperheater removed for service

Figure 8: Competitor desuperheaters; the left one is used under normal conditions

This desuperheater is installed too close to the first downstream bend, and it also sprays part of the cooling water directly onto the pipe wall. Thermographic images are expected to show the thermal stress of the pipe caused by streaks of water that move around on the pipe wall. The method is also expected to give a picture of the poorly performing evaporation process in the pipe.

Indeed, the thermographic images in Figures 9, 10, and 11 reveal that the water spray covers only a part of the cross section of the pipe. The temperature of the warmer zone is close to the upstream temperature, and the temperature of the colder zone is close to the saturation point.

Figure 9: Thermographic image of steam pipe 4.6 to 6.6 feet (1.4 to 2.0 meters) after the DA-M desuperheater: the liner gives the yellow color

Figure 10: Thermographic image of steam pipe 6.6 to 8.5 feet (2.0 to 2.6 meters) after the DA-M desuperheater: the liner gives the yellow color

Figure 11: Thermographic image of steam pipe 8.5 to 10.5 feet (2.6 to 3.2 meters) after the DA-M desuperheater: the liner gives the yellow color

According to the field test of the competitor's desuperheating application, the atomization and evaporation of the injected water is ineffective – a great deal of the water is sprayed directly onto the pipe wall on one side, and no water is sprayed into the opposite half of the pipe's cross section.

The thermographic camera clearly reveals a distinct problem with a poorly performing desuperheater. The overspray on the hot piping can lead to cracking, and the uneven temperature distribution will lead to water dropout and poor process control.

The graphs in Figures 12 and 13 show a comparison of the temperature data taken during these two tests, indicating the superior performance of the DA-M design with complete evaporation occurring in 45 milliseconds.

Figure 12: Evaporation time downstream from a CCI desuperheater based on temperatures from thermographic images

Figure 13: Temperature variation downstream of the poorly performing desuperheater based on thermopraphic images

Indications

CCI Sweden's field tests successfully evaluated atomization performance, calculating droplet distribution after secondary atomization using downstream temperature distribution. Through these studies, CCI Sweden proved that ring-type desuperheaters with a liner, such as CCI's DA-M desuperheater, are significantly better than simple probe-style desuperheaters.

This testing also shows that the thermographic images offer a valuable tool when addressing desuperheater applications that are performing poorly. By clearly showing the causes of the problems, CCI will be better equipped to offer the precise solution for its customers. During the paper mill's next outage, the probe-style desuperheater will be replaced with CCI's DA-M design with a liner.

Published in SOLUTIONS Spring 2004