Case Study - Upgrade from F9-class to Gore HEPA E12-class filters: A success story from Kuwait Oil Company

Use of single-stage Gore HEPA E12 filters on compressor-driven gas turbines eliminates sand ingression on first-stage nozzles, resulting in significant savings

The timeline of upgrade by Kuwait Oil Company from F9-class filters to Gore HEPA E12-class filters, at the Ahmadi oil fields.
The timeline of upgrade by Kuwait Oil Company from F9-class filters to Gore HEPA E12-class filters, at the Ahmadi oil fields.

Kuwait Oil Company (KOC) operated stationary aero-derivative gas turbines for various applications in a desert environment. At the Ahmadi oil fields, gas turbines were used to drive four compressors that are critical to oil production operations. There was no back-up; therefore, these compressors cannot afford to shut down for events such as cleaning, maintenance, or repair.

As things stood, the gas turbines had to be shut down every one to three months for compressor washing due to contamination. Conventional F9 turbine filters were used, which had to be replaced every 10 to 13 months. This resulted in reduced performance coupled with high overhaul costs (maintenance/filter changes).

It was not realistic to replace the damaged turbine parts on a regular basis as it involves high repair cost and significant production loss due to unit downtime. The gas turbine and filter house OEM had recommended major modification to the gas turbine air intake filtration system to resolve the sand ingression issue. KOC determined that the OEM’s solution was also unrealistic and very expensive due to the high capital investment and production loss.

To overcome these challenges, KOC sought an alternative solution. They contacted Gore to explore whether an upgrade to Gore’s HEPA E12-class single stage pulse-cleanable filters, in a direct one-to-one replacement of the existing F9-class filters without any modification to the existing set-up, would resolve the sand ingression issue.

Findings at site and during GG dismantling

Based on the findings at site and during GG dismantling, it was clear that a large quantity of sand was getting into the flow stream and contaminating the gas generator internals. The main source of the sand ingress could be classified into two main contributory factors: (1) Installation errors leading to bypass of filtration stage. (2) Design insufficiency of the F9 filters to provide clean, dust- and sand-free air for combustion.

To attend to point 1, KOC coordinated with the turbine OEM and rectified all bypasses in the filter house. The corrective action helped to reduce the visible dust on the downstream side of the filter house. Contamination on the stage 1 nozzles, however, was not reduced.

Contamination in the form of fused silica in the combustion section raised some serious questions. OEM identified the quality of air supplied to the gas turbine as one of the major causes of contamination on the compressor and gas turbine components. To confirm the effects of sand contamination on the gas turbine, KOC engaged a third-party consultant to verify the findings independently.

The third party report is classified into following categories: (1) Failure of the gas generator due to sand ingress into the main combustion flow path – The sand in the flow stream gets exposed to high firing temperatures, resulting into change of state from sand to molten silica, which then deposit onto the first stage nozzles and further downstream onto 1st stage rotor blades. The deposition severely compromises the efficiency of the cooling holes leading to further secondary damage on the blades thereby exposing the hot section to significant risk of catastrophic damage/blade failure. Seven times the gas turbines were removed from service on OEM recommendation due to nozzles/blade sand contamination. (2) The gas generator by design utilises external filtered air to supplement internal bleed air extraction for bearing sump pressurisation. Two units were removed from service because of bearing failures due to silica contamination in oil, one possible source of contamination is through the external bearing sump pressurisation air. The effected units were removed from service and sent for complete strip and repair.

In both the above cases, internal hardware of GG was effected, Failure RCA and repair was carried out at OEM repair facility. A total of 102,487 hours of serviceable life was lost due to premature damage on the machines.

Failure mechanism leading to accumulation on nozzles/blades

It was clear from the Borescope pictures of HPT blades and vanes that deposition is evidently higher for stage 1 components. Over the years, different non-TBC coatings have been applied to LM2500 HPT blades and vanes. The use of TBC coating on stage 1 vanes was introduced with the LM2500+G4/PGT25+G4 engine. It is believed that the coating type is not a key parameter that would significantly increase, or decrease the rate of sand deposition. A confirmation of the latter is a comparison of stage 1 and 2 blades, which are suspected to have the same type of coating and present significantly different signs of deposition. Based on the present industry knowledge, coatings that prevent sand deposition do not currently exist.

Evidence that there are traces of sand particles greater than one micron inside the inlet duct (clean side) was presented based on sand particle size measurements. When sand particles with size distribution as found after site visit, the large particles cannot follow streamlines and are likely to impact airfoils. Particle size tends to decrease as solid particles travel through axial compressor. It has been concluded that particles greater than one micron are likely going to impact HPT stage 1 vanes and stick to airfoils.

It has also been concluded that there may be particles slightly larger than one micron that can make it past stage one vanes and impact stage one blades. Particles smaller than one micron are likely to follow the streamlines and are less likely to impact stage 2 vanes and blades. Even if impacts occur at stage 2, particle temperature is lower and deposition is less likely to occur.

Turbine flame temperature for the PGT25+G4 engine is expected to be higher than 1,600°C, at base load. Turbine inlet and stage 2 temperatures are expected to be around 1,300°C and 1,000°C, respectively. Based on understanding from theoretical concepts, sand particles are less likely to stick to surface when temperature is below 1,000°C. It was determined that main contributors to the sand deposition and engine failures include the presence of sand particles of size bigger than one micron, temperature of sand particles at turbine inlet temperature, and temperature difference of molten sand particles and HPT stage 1 components.

Based on the evidence and analysis conducted by KOC in association with a leading gas turbine consultancy service provider, it was determined that the main contributors to deposition and engine failures include the presence of sand particles of a size bigger than one micron, temperature of sand particles at turbine inlet temperature (~1,300°C), and temperature difference of sand particles and HPT stage 1 components. The F9 filters cannot remove all particles below 4μm. Since it is found that particles from 1μm to 4μm can lead to sand deposition, the current filtration system is not adequate. A filtration system with higher particle removal efficiency for the size range of 1μm to 4μm is needed.

Upgrading to GORE Turbine Filters with E12 hydrophobic HEPA technology has provided enormous value to rotating equipment users.

Return on investment

After 15 months of operation with the GORE Turbine Filters, the basic intent of the KOC upgrade from F9-class filters to Gore HEPA E12-class filters was achieved. Molten sand ingression was eliminated in the 1st stage nozzle and rotor blades in the hot section of the gas turbine. The benefits of this are: eliminating premature failure of gas turbine parts and achieving OEM scheduled service life of 25,000EOH; significant savings on gas turbine part premature failure repair cost of $21.1mn; and increase in production due to continuous availability of gas turbine trains.

Additional benefits from using Gore HEPA E12-class filters are: compressor efficiency increased by 1.5%; savings of an average of 3.05% of fuel costs due to more efficient fuel consumption; lower and more stable differential pressure over filter lifetime compared to F9-class filters; and longer filter service life – 18 months for Gore HEPA E12-class filters compared with 12 months for F9-class filters.

Considering these field-proven performance benefits, the higher price of the Gore HEPA E12-class filters compared to the F9-class filters was justified within a few days of operation. The Gore HEPA E12-class filters have demonstrated excellent performance in the KOC mechanical drive (gas generator) application.

For the latest refining and petrochemical industry related videos, subscribe to our YouTube page.

For all the latest refining and petrochemical news from the Middle East countries, follow us on Twitter and LinkedIn, like us on Facebook.

You may also like

2019 RPME Top 30 EPC Contractors: Hyundai Engineering Co ranked No. 26
$1.39bn: For building the regasification plant
2019 RPME Top 30 EPC Contractors: MMEC Mannesmann ranked No. 24
Fifty years of experience in implementing projects
2019 RPME Top 30 EPC Contractors: Nuberg EPC ranked No. 23
$53mn: Chlor alkali and calcium chloride projects