Proceedings of GT2005 ASME Turbo Expo 2005: Power for Land, Sea and Air June 6-9, 2005, Reno-Tahoe, Nevada, USA GT2005-68701 AXIAL COMPRESSOR DETERIORATION CAUSED BY SALTWATER INGESTION Elisabet Syverud Olaf Brekke NTNU Royal Norwegian Navy Norwegian University of Science and Technology Bergen, Norway Department of Energy and Process Engineering N-7491 Trondheim, Norway Lars E. Bakken NTNU, Norwegian University of Science and Technology, Department of Energy and Process Engineering, N-7491 Trondheim, Norway ABSTRACT From operating experience and literature the first stages of Gas turbine performance deterioration can be a major an axial compressor are expected to be the ones most heavily economic factor. An example is within offshore installations fouled. Deposits will have different characteristics depending where a degradation of gas turbine performance can mean a on the nature of the fouling. Dry particles in dry atmospheres reduction of oil and gas production. are likely to deposit in different areas compared to sticky matters This paper describes the test results from a series of and oily compounds. At high inlet humidity, the drop in static accelerated deterioration tests on a GE J85-13 jet engine. The pressure during acceleration will increase the dust adhesion on axial compressor was deteriorated by spraying atomized the blades [3,4]. droplets of saltwater into the engine intake. To understand and reveal the degradation and restoration The paper also presents the overall engine performance mechanisms related to recoverable compressor deterioration a deterioration as well as deteriorated stage characteristics. The systematic test campaign has been performed on a GE J85-13 jet results of laboratory analysis of the salt deposits are presented, engine. The overall performance of the gas turbine, and providing insight into the increased surface roughness and the compressor stage characteristics were analyzed in detail. The deposit thickness and distribution. tests reflect actual performance deterioration as found in The test data show good agreement with published stage offshore applications and provide valuable information characteristics and give valuable information regarding stage- regarding overall and stage-by-stage compressor performance by-stage performance deterioration. deterioration in a real engine. The tests were part of a larger online water wash test Keywords: Axial compressor, salt, roughness, stage program. The results from this program are reported by Syverud characteristics, GE J85-13 and Bakken [5], and that paper should be seen in context with the present work. INTRODUCTION Compressor fouling causes 70-85% of the performance loss NOMENCLATURE due to deterioration in gas turbines [1]. The degradation rate will A Effective flow area, [m2] be affected by site-specific conditions of the fouling and by cp Specific heat at constant pressure [kJ/kgK] engine operating conditions. A thorough review of engine C Absolute air velocity, [m/s] performance deterioration mechanisms and modeling is given by CIT Compressor inlet temperature, [K] Kurz and Brun [2]. CIP Compressor inlet pressure, [kPa] 1 Copyright © 2005 by ASME GE General Electric service in 1960. More than 6000 engines are still in service IGV Inlet guide vanes worldwide. The Royal Norwegian Air Force (RNoAF) F-5 ISO Int. Organization for Standardization (*) Freedom Fighters have two GE J85-13 engines for propulsion. k Grain size of salt crystals, [µm] Because these aircrafts have been replaced by the F-16 aircraft, ks Equivalent sand roughness, [µm] one GE J85-13 engine has been made available to the Norwegian m Mass flow rate, [kg/s] University of Science and Technology (NTNU) for use in this MW Molecular weight [kg/kmole] project. Engine testing was carried out at the RNoAF’s test N Shaft speed, [rpm] facilities at Kjeller, Norway. Kjeller is located inland, ~25 km east NACA National Advisory Committee for of Oslo at an altitude of 119 mabove sea level. Aeronautics The GE J85-13 is a compact, light-weight, single-spool NASA National Aeronautics and Space turbojet engine. It has an eight-stage axial-flow compressor that Administration is directly coupled to a two-stage turbine. The exhaust nozzle P Pressure, [kPa] has a variable throat area, and the engine is equipped with an PS Pressure side (concave) of blades/ vanes afterburner. The GE J85-13 is rated to deliver a minimum thrust of R Gas constant, [kJ/kgK] 2720 lbs (12.1 kN) at 100% shaft speed without the afterburner R0 Universal gas constant, [kJ/kgK] and a minimum thrust of 4080 lbs (18.1 kN) with full afterburner. Re Reynolds number The eight-stage compressor has a pressure ratio of 6.5:1. R.H. Relative humidity, [%] First- and second-stage rotor blades are coated for corrosion RNoAF Royal Norwegian Air Force protection. To handle off-design operation, the compressor has RTD Resistance temperature detector variable inlet guide vanes (IGV) and bleed-off-valves at stages SS Suction side (convex)of blades and vanes 3-5, as well as a variable exhaust nozzle. The variable geometry T Temperature, [K] is controlled by the throttle angle, but the timing is ambient U Blade velocity, [m/s] temperature biased. At ISO conditions, IGV will be at maximum V Relative air velocity, [m/s] deflection and bleed-off-valves will be fully open below ~81% war Specific humidity, [kg H2O/kg dry air] corrected speed and fully closed at >96% corrected speed. At x Molar fraction of component ambient temperatures above ISO conditions, the bleed-off- α Absolute air angle, [deg] valves will close at lower speed settings. The nozzle is β Relative air angle, [deg] controlled by the throttle as long as the engine is running below ρ Air density, [kg/m3] the exhaust temperature limit. When the maximum exhaust η Polytropic efficiency temperature is reached, the exhaust nozzle will increase the flow. ν Kinematic viscosity, [m2/s] This reversal in the exhaust nozzle schedule occurs close to the γ Ratio of specific heats maximum throttle angle. T 2 2 Ψ =cp∆Tt/Utip Stage work coefficient, [(kJ/kg)/(m/s) ] Bleed air is also used for internal cooling of the afterburner 2 Φ =Ca/Utip Flow coefficient actuators, combustor liner, and turbine guide vanes. Bleed air Subscripts for anti-ice is taken at the compressor discharge. The anti-ice air a Axial velocity component valve was kept closed during all tests associated with this amb Ambient condition project. mix Mixture (humid air) s Static condition Engine model data t Total condition Detailed knowledge of compressor stage geometry is tip Tip available through several sources providing a good basis for VMD Volume median diameter developing engine simulation models. w Tangential velocity component NASA carried out several tests on the GE J85-13 in the late 2.1, 2.2,... 2.7 Compressor stage 1, 2,... to stage 7 1960’s and throughout the 1970’s. Hager [6] analyzed internal 3 Compressor discharge flow of the GE J85-13 multiflow compressor and reported 5 Turbine discharge compressor stage temperature and pressure profiles. Milner and Wenzel [7] studied the effect on compressor performance from TEST FACILITIES AND ENGINE DESCRIPTION clean and distorted inlet flow, and their work includes individual The General Electric (GE) J85-13 engine has been a compressor stage characteristics. Compressor characteristics successful engine in different applications since it first entered are also given by Willoh and Seldner [8], whereas engine geometry is given by Tesch and Steenken [9]. * ISO reference conditions assume an ambient temperature of 288.15 K, a barometric pressure of 101.325 kPa and a relative humidity of 60%. 2 Copyright © 2005 by ASME Initial engine condition temperature. CIT was measured using four sensors located at The engine used in this test program was preserved and the engine inlet screen. stored after serving ~300 hours onboard an aircraft. The engine All instruments were calibrated prior to the test program was not overhauled prior to testing. and the measurement uncertainties were calculated based on To avoid disturbances to the airflow attributable to loose methods given in the ASME Performance Test Codes [11,12]. coating, the coating on the second-stage rotor blades was To reduce the data scatter, a minimum of 60 data points (2Hz removed by hand prior to the testing, using emery paper grade sampling rate) were collected at each setting. The average of 320. The compressor inlet, inlet guide vanes, first-stage rotor, these readings was taken as the steady-state data point. and stator were cleaned by hand using solvent, while the upper half of the casing, including stator vanes from stages 1-7 were ENGINE DEGRADATION METHOD AND EQUIPMENT cleaned using an ultrasonic cleaning bath. In addition, online Accelerated engine deterioration was done through the compressor cleaning was performed prior to the testing using ingestion of atomized saltwater.A similar method has been used water at a flow of 13.2 l/min, for a 30 second period at idle, 70, 90, by the United States Navy [13]. The saltwater ingestion system 95 and 100% corrected engine speed. Details of the online water is illustrated in Fig. 2. Saltwater was atomized in front of the wash system are given by Syverud and Bakken [5]. compressor inlet using four atomizing nozzles installed on a bracket on the online water wash manifold positioned 0.77 m in ENGINE INSTRUMENTATION front of the compressor inlet. This ensures complete atomization Additional sensors were installed to provide more of the saltwater prior to the inlet guide vanes. The saltwater was information than is available from standard, test-cell gravity fed from a storage tank while air pressure was taken from instrumentation. The engine instrumentation is shown in Fig. 1. the pressurized air supply in the test-cell. Figure 1 GE J85-13 engine instrumentation Figure 2 Salt ingestion system The temperatures at stages 1, 3, 4, 6 and 8 were measured using a single resistance temperature detector (RTD) at each The droplet diameter of the saltwater spray was measured µ stator row.