یک روش برای تعیین هزینه های اقتصادی از رسوب کمپرسور توربین گاز

A method to determine the economic cost of gas turbine compressor fouling was developed. The method is based on the calculation of the influence of the drop of compressor performances (isentropic efficiency, pressure ratio and air mass flow) on Heat Rate and full load power of the gas turbine. A statistical procedure to quantify the drop of compressor performances with fouling and to obtain ambient correction factors for the compressor was also developed. Finally, this method was applied to a real gas turbine in order to check its validity. It was demonstrated that the application of the correction factors obtained worked well. The method was used to check the evolution of the compressor efficiency over time and the effects of off-line washings on compressor performances. It could be concluded that the loss of maximum power is the most important factor to take into account to schedule off-line washings.

During recent years, the Department of Thermal Engineering of the Universidad Politécnica de Madrid has conducted several studies about Combined Cycles. Some of them are related to the heat recovery steam generators [1] and [2] or to the thermoeconomic optimization of combined cycle using genetic algorithms [3]. Continuing with this research line, it was decided to carry out a study about fouling of gas turbine compressors, which is currently being studied by a number of authors like Sánchez et al. [4] or Aretakis et al. [5]. Like any other machine, the performance of gas turbines is affected by wear and tear over time. Kurtz and Brun [6] list the mechanisms which cause the degradation of gas turbines. Some of these mechanisms are commonly referred to as “non-recoverable”, especially those related to hot gas path components, which might require maintenance such as welding or eventually the replacement of a damaged component. Compressor fouling is said to be “recoverable” by means of light maintenance such as washing. It is caused mainly by particle deposition on the blades and annulus surfaces of the compressor due to the ingestion of dust mixed with the air and not blocked by the inlet filter. Compressor blade fouling reduces the flow capacity, the isentropic efficiency and the pressure ratio. Zwebek and Pilidis [7] explained that fouling affects the thermal efficiency and output power of the engine. Lakshminarasimha et al. [8] said that fouling is one of the most common causes of performance reduction encountered by gas turbines. According to Vigueras [9], the economic production of a 240 MW gas turbine is, on average, 106 million USD per year and, if the fouling affects 1% the compressor pressure ratio, the cost of fouling is about 6.25 million per year. The study performed by Sánchez et al. [4] concluded that a good scheduling maintenance tasks could save electrical companies 200.000 €/year/gas turbine. So it is essential to adopt a correct strategy to carry out the washing processes, which may be done on-line or off-line. Gülen et al. [10] described the different methods employed to clean compressors. In an on-line wash, distilled water is injected into the compressor during normal operation in order to remove deposits. However, complete performance recovery can only be achieved by an off-line wash, where distilled water mixed with a special detergent is sprayed into the gas turbine which is rotated by the starter. Fig. 1, similar to a Figure used by Naga and Achutha [11], shows the qualitative effects of performing off-line or on-line washings on compressor efficiency. Different methods for scheduling compressor washings have been proposed by Sánchez et al. [4], Aretakis et al. [5], Naga and Achutha [11] or Hovland and Antoine [12].This paper provides a simply and accurate method to calculate the economic cost of fouling which can be used in methods for scheduling compressor washings like the aforementioned. In order to achieve this goal, the paper is divided into 3 different parts. Section 2 explains the two effects that compressor fouling may cause: over-consumption (Heat Rate increase) and drop of full load power. This section also shows the way used to quantify economically these effects. Section 3 shows the method used to find the influence of the compressor performances (CPs) (isentropic efficiency, pressure ratio and air mass flow) on Heat Rate and maximum power. Sections 4 and 5 explain the way used to know the CPs drop due to fouling. CPs are significantly influenced by environment. The papers of De Sa and Zubaidy [13] and Meher-Homji and Bromley [14] explain the influence of environment. It is not a simple task to compare CPs between clean and dirty compressor under different ambient conditions. In order to solve this problem, correction factors are defined to refer gas turbine's features to ISO conditions (Ta = 15 °C, pa = 1atm). Original Equipment Manufacturers (OEMs) usually provide correction factors for the whole gas turbine cycle but not for the isolated compressor. Finally, the whole method to calculate the economic cost of fouling was applied to a real gas turbine in Section 6. This allows checking the aforementioned method and to carrying out a study that quantified the economic cost of the fouling.

A method to calculate the economic cost of compressor fouling was developed. This cost is due to an increase in Heat Rate and a drop in full load power, which are consequences of CPs drops. After applying this method to a real gas turbine, it was obtained the main conclusion: over-consumption might be considered negligible in comparison to the cost of full load power loss. This is important to schedule off-line washings because when the gas turbine is running at its maximum capacity, the economic cost of fouling might be higher than the opportunity cost of performing an off-line washing and in that case it would be advisable to produce energy through another way and to carry out the washing. Furthermore, the method explained in this paper allows the compressor performances to be checked regardless of influence of ambient conditions. This is useful for performing different studies into the mechanism of fouling and the influences of factors like the time of year, the environment, the quality of filters… This method also allows checking the effect and the success of a washing (Fig. 6). It was checked that the effects of washings are not always the same, so this should be something to be analyzed in future works. Finally, the method explained in this work presents some weaknesses. First, it is necessary to dispose of high accuracy sensors which are not always available in every gas turbine. Second, it would be convenient to operate the gas turbine at different loads after each off-line washing in order to provide data running in clean conditions and this is not always possible.