Failure Analysis of Once-Through Steam Generator (OTSG) Tube

A cogeneration plant produces energy in the form of both electricity and steam [1]. About one-third of the input energy is converted to useful electrical power in a conventional power plant. At the same time, two-thirds are thrown away as waste heat to rivers or the atmosphere [2]. Recently, cogeneration in the form of municipal power and district steam heating has been gaining immense popularity to fulfill their objectives like energy-saving, environmentally friendly, and compliance with the safety and regulatory boards. Briefly, wasted heat from electricity generation generates steam, which is sold to customers like the auto industry and other different businesses.

A power plant has many crucial components, like a once-through steam generator (OTSG) collaborating to generate electricity [3]. The once-through steam generator is a continuous tube in which preheating, evaporation, and superheating of the working fluid take place consecutively. Water enters at one end of the OTSG through the inlet header and exits the other end of the OTSG as superheated steam through the outlet header. It has good static performance, stable steam pressure, good operation maneuverability, fast power in rise and fall, and superheated steam is produced to improve thermal efficiency [4]. Hence, detecting OTSG tube leakage is vital, as approximately 60% of boiler outages are due to tube leaks [5]. The failure of crucial components in power plants requires prevention [6]. Developing the OTSG tube failure detection system is challenging for modern power plants’ safe and reliable operation [7]. Many steam generator tubes must be fixed and removed from service or repaired yearly [8]. This widespread damage has been caused by many diverse degradation mechanisms, some of which are difficult to detect and predict. These ruptures have caused complex plant transients, which have not always been easy for power plant operators to control [9]. OTSG tube failure used in an oil company caused unexpected severe degradation at welds due to erosion–corrosion [10].

A growing share of electricity comes from distributed generation sources like wind and solar [11]. The power output of renewable sources is not constant and fluctuates depending on weather conditions. Power plants should provide flexible capabilities to meet the ever-changing demand to stabilize the grid frequency, and substations withstand the transient behavior of the generated power [12]. In addition, to reducing the mismatch between the demand and supply during the day and night and different peak and off-peak seasons, the start and stop cycling of the power plants are increased for the daily start and stop operations [13,14]. During the startup of power plants, thermal stresses in components of the OTSG are caused by exposure to high-temperature gas and steam. These thermal stresses decrease the life of the components like tubes [15]. A reduction of tube consumption is required to extend the OTSG design lifetime.

This paper involves data from a real-world cogeneration power plant. It will comprehensively analyze the leakage on the bottom wall tubes of an OTSG in a cogeneration power plant. This unit of OTSG produces 150,000 lb/h steam. It gets heat from a 55 MW combustion turbine generator (CTG). The leakage was found during a hydrostatic test as an annual maintenance activity. The stress analysis of the internal working pressure in the tubes, the chemical composition of the feed water and failed tubes, including the deposits around the crack, and the analysis of the microstructural morphology will be reported. Furthermore, the analysis related to the failure of boiler tubes in steam power plants can be a reference and guideline to perform the appropriate precautions according to the proposed measures to consider some practical engineering activities.

The natural gas-fired cogeneration plant generates 300,000 lb/h of steam and 110 MW of electricity. The primary function of the main steam system of the plant is to provide high temperature (HT) and low temperature (LT) steam to the distribution system. The secondary function is to provide HT steam to the steam turbine generator (STG), CTG inlet air heating system, deaerator pegging steam, and office building heating system. This main steam system receives HT steam from two OTSGs and a package boiler. The steam exiting the OTSG is desuperheated to LT steam conditions using a desuperheater. OTSG is designed per ASME boiler and pressure vessel code (BPVC) section I [16]. The design pressure and temperature of the OTSG are 366 psia and 965 °F, respectively, to produce 150,000 lb/h superheated steam. All tubing is furnished to the chemical requirements of ASME Specification SB 407 and UNS N08800 alloy with continuous fins to increase the heat exchanger surface area. The OTSG has an integral gas inlet transition duct section and an integral exhaust stack. The gas turbine connects to the inlet duct using an expansion joint. Around the perimeter of the tube bank, end and side seals are provided between the tube bank and the wall liner at several elevations to prevent turbine exhaust gas from bypassing the tubes. The entrance to each tube is just downstream of the feedwater headers and contains an orifice, which ensures even flow distribution of the feedwater to all tubes. Figure 1 shows the OTSG at the plant and the location of the faulted tube.

OTSG does not have defined economizer, evaporator, or superheater sections. The single control point for the OTSG is the feedwater control valve; actuation depends on predefined operating conditions set through the plant’s distributed control system (DCS). The DCS is connected to a feed-forward and feedback control loop, which monitors the transients in gas turbine load and outlet steam conditions. The control loops set the feedwater flow to a predicted value based on the turbine exhaust temperature to produce steady-state superheated steam conditions.

At the onset of steam production, produced steam is vented to the atmosphere or cooled using an attemperator before admission to the plant steam piping or turbine condenser bypass systems. There are steam desuperheating stations to maintain proper downstream conditions of the steam to the main steam line. A portion of the HT steam from each unit is desuperheated to the LT steam condition using the desuperheater spray water from the feedwater system.

Feedwater entering the OTSG is converted to steam. During operation, the feedwater must be of sufficient quality to ensure that no scaling occurs inside the tubing and that the purity of the steam output is suitable for the process. pH and water and steam cation conductivity are continuously monitored to confirm that the water quality is within the water chemistry control limit and guidelines [16–20], EPRI [21], and OTSG manufacturer’s specifications. Because there is no separation of phases in OTSG, the feed water must be of very high purity to meet the OTSG operation requirements. The condensate polisher is used for satisfactory operation for OTSG because feedwater purity requirements are incredibly stringent. Continuous water samples are drawn from the OTSG steam-condensate-feedwater cycle and analyzed to monitor the impurities in the system. OTSG was installed in the cogeneration plant and commissioned as per ASME code requirements. After four years of installation of the OTSG, the leakage in the tube was noticed. The leak occurred at a finned section of a straight tube. A segment was removed from the lower pass at the bottom of the OTSG for examination. A drawing showing the tube layout in OTSG and the tube failure location is provided in Fig. 1. The steam exiting the OTSG has a temperature of 655 °F at a pressure of 305 psig. The gas side temperature for this section of the generator is 840 °F. The tube from which the failed section was removed was produced from 1 to 1/4 in. diameter material having a minimum wall thickness (MWT) of 0.054 in. The tubing was specified as Inconel 800. It was furnished to the requirements of ASME Specification SB 407.

As part of this metallurgical evaluation, the failed tube segment was initially examined visually, and the conditions were documented. Figure 2 shows the external surface of the leaked tube section. The external surface of the tubing after the removal of the fins at the suspected failure site is shown in Fig. 3. It shows that around the crack, there are red deposits on the pipe. A close-up photograph of the external surface shows the presence of a transverse crack, as shown in Fig. 4. The presence of red deposits surrounding the crack is visible. It is evidence of deposition due to the evaporation of steam or superheated water through the crack.

The tube segment was split longitudinally to allow the inner diameter (ID) surfaces to be examined. Figures 5 and 6 are photographs of the inner surface of the leaked tube. A layer of red deposit on top of a silver scale covered most of the internal surface. The deposited layer was loosely bound. The nature and pattern suggested it was deposited on the surface. No thick scale build-up from in situ oxidation was noticed. Evidence of cracking was observed where a streak of red deposit had not formed. It is because the steam or superheated water was leaking through the crack with a strong force. Portions of the tubing were cleaned using a glass bead blasting technique so that the underlying contour of the metal surface could be examined. Figure 7(a) shows the photograph of the external surface in the failed area. Figure 7(b) is the zoomed-in image of the cleaned external surface where there is a crack. Figure 8 is the close-up photograph of the internal surface at a through-wall crack after cleaning. The tube wall thickness was a maximum of 0.081 in., higher than the design minimum requirement of 0.054 in..

A three- and one-half-inch long test section was removed from a representative tube area to determine the deposit-weight-density (DWD) value. The glass bead blasting method was used to determine the DWD value for the test section. Micrometer measurements were taken before and after the deposit removal step was used to measure the deposit thickness of the internal surface. The DWD value was a maximum of 5.1 g/ft². The maximum internal deposit thickness was 0.002 in..

Scanning electron microscope-energy-dispersive X-ray analysis (SEM-EDXA) was used to determine the elemental composition of the internal deposits. The gray and red deposits were analyzed separately, as shown in Figs. 9 and 10, respectively. The results are listed in Tables 1 and 2 for gray and red areas, respectively. The internal gray deposits primarily comprised iron, chromium, and nickel compounds. There were also fewer amounts of sodium, silicon, aluminum, potassium, and calcium species. The top layer of the red deposit was primarily iron oxide.

Nickel alloys such as Inconel Alloy 800 are susceptible to stress-corrosion cracking (SCC) when exposed to high temperatures in caustic environments (sodium/potassium hydroxides). Minor amounts of sodium and potassium species were detected on the internal surface using SEM-EDXA in the failed tube. Caustic-induced SCC typically produces intergranular cracking. In the subject failed tube, the caustic was concentrated by evaporation when water flashed to steam in the failure locations. Feedwater quality and steam purity are interrelated in that feedwater quality affects steam purity. Table 3 provides the boiler feedwater specifications. Sodium and silica are the constituents of the feedwater. The recommended pH value for the OTSG tube is based on achieving improved condensate polisher performance. The conductivity and pH of the sample collected during plant operation are measured at 77 °F. As per feed water chemistry analysis from the OTSG feed water sample collected from piping at the OTSG inlet and on the steam piping at the exit of the OTSG shows that the water quality controlling parameters are within the limit. During operation, the leaking pressure gauges were identified. It is due to a defect in the resin in the condensate polisher. High amine usage was recorded to control the pH of feedwater strictly.

A portion of the tube was analyzed using optical emission spectroscopy (OES) to determine the alloy composition. Chemical testing was performed on an optical emission spectroscopy with the ASTM-E1086. Carbon and sulfur content was determined by combustion testing in accordance with ASTM-E1019. The results shown in Table 4 confirmed that the tubing was fabricated from a material consistent with chemical specifications for Inconel Alloy 800 (UNS N08800).

Transverse sections were removed from the failed regions of the steam hose and prepared for metallographic inspection. Figures 11–14 show the tube’s inner surface after cleaning and magnifying at 50, 100, 200, and 500 times, respectively. Clearly, a crack is seen, which has no defined shape. Microstructural examination indicated that the failures resulted from cracking that was initiated on the internal surface. The cracks were branched, disjointed, and intergranular. The cracking damage appeared to be characteristic of SCC. Figure 15 shows the mid-wall microstructure of the OTSG tube. Away from the cracking, the microstructure appeared typical for Inconel 800 in the as-fabricated condition.

The transverse crack path in the subject tube revealed that the stresses responsible for SCC were axially oriented; the tube was pulled at its ends, and the stresses were due to bending or thermal expansion and contraction. During the OTSG operation, the operating pressure, temperature, and feed water flow contributed to applied stresses in the failed tubing. It is typical for residual and applied stresses to contribute to the formation of SCC, such as at the welds of a load-bearing welded attachment. The leak location of the failed tube was a straight tube between a U-bend and a reducing section of the tube. Residual stress developed because of improper heat treatment during manufacturing. Improper heat treatment or heat from welding could be an influencing factor for intergranular corrosion in OTSG tube failure. The improper heat treatment caused the precipitation of specific alloy components at the grain boundary during the steam or superheated steam flow. There was no evidence that the corrosion in the failed tube was due to the sensitization of the base metal in heat-affected zones. The evidence reported in this tube failure laboratory examination (Fig. 13 and Table 1) suggests that SCC is associated with an anodic peak in the active region of potentials. Anodic dissolution is the governing mechanism for caustic crack.

A tube leak was identified at a superheater tube in OTSG. A transverse crack was observed on the external surface of the failed area located at the bottom section of the OTSG. The red color around the crack was visible. It is deposited due to the evaporation of steam or superheated water around the crack. In the failed tube’s internal surface, evidence of cracking was observed where a streak of red deposit had not formed. It is because the steam or superheated water was leaking through the crack with force. There is no evidence of overheating damage around the crack. In the subject failed tube, sodium and potassium species were detected on the internal surface. Caustic was concentrated by evaporation when water flashed to steam in the said location. Caustic comes from the high amine, which is required to control the pH of OTSG feedwater. The corrosion attack occurred when concentrates from water as steam were produced in this generating tube.

A metallurgical evaluation confirmed that the cracks were branched, disjointed, and intergranular. The SCC caused the OTSG tube failure. The transverse crack path in the subject tube revealed that the stresses responsible for SCC were axially oriented; hence the tube was pulled at its ends, and the stresses were due to bending or thermal expansion and contraction. Operating conditions pressure, temperature, and feed water flow contributed to applied stresses in the tubing.

The analysis identified that the OTSG feed water quality was responsible for the SCC in the OTSG tube. The makeup water, condensate polisher performance, and chemical feed to control the water quality parameter influenced the corrosion. To meet the changing demand, rapid pressure change during the OTSG startup from idle and the residual and applied stresses created corrosion in the failed tube. Creep-fatigue loading will be considered in future tube failure analysis as a factor of intergranular cracking of the tubes.

The authors acknowledge the technical support given by the operation and maintenance team of the REO cogeneration plant and Asset Management Department. The authors also acknowledge the support given by Dave Bolan and Stephen Serkaian of the Lansing Board of Water and Light.

There are no conflicts of interest. This article does not include research in which human participants were involved. Informed consent not applicable. This article does not include any research in which animal participants were involved.

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.