1

Graeme Turnbull – AAF

Erwan Clément - Donaldson

Tord Ekberg – Camfil Farr

WG3; To extend the ultimate life and repair interval for key hot section components by

30%

WG4; 25,000 hours of gas turbine operation without intervention.

It is proposed research be undertaken to establish a framework of metrics to correlate

best practice filtration in relation to compressor efficiency. The ultimate aim of the

research shall be gas turbine performance enhancement and component life extension

in line with the vision statements.

This paper shall discuss the level of filtration required to meet existing OEM (Gas

Turbine Original Equipment Manufacturer) specifications, existing filtration international

test standards and the commercial and technical benefits available to operators by

applying enhanced Hepa H Class filtration technology to their gas turbine fleet to

significantly reduce fouling of the compressor blades and consequential power loss.

The compressor of a gas turbine consumes a significant amount of energy during

operation; consequently, the efficiency of the compressor is very important to maintain

optimum performance and has a huge impact on the machine thermal efficiency, power

output and its long term component health.

Engine performance and component life should be considered as a function of the total

mass of contaminant ingested which is directly influenced by the type of atmospheric

and industrial environment; these deposits decrease the air flow performance of the inlet

compressor due to degradation in blade shape and surface finish. Ultimately the overall

performance of the turbine is greatly affected.

2

stringent & high quantities of dust, aerosols and water continue to pass through the

filters every second and deposit on the blades of most engines in use today.

More normally associated with micro electronics production and laboratory / hospital

protection, Hepa filtration provides particle removal efficiencies of up to 1000 times

greater at the critical sub-micron sizes than achieved by traditional reverse pulse and

The primary benefits derived from enhanced filtration technology include:-

- Greater machine availability (%)

- Consistent and higher power output

- Increased fuel efficiency

- Longer hot end component life

- Reduced/negate water wash process

- Improved reliability

- Lower emissions

The potential commercial upside considers:-

- Increased plant revenue

- Greater production yield (i.e. Oil & Gas, Steam)

- Lower labour and fuel costs

- Lower component costs

- Greener technology use

It is a fact that the performance of the air filtration system has a huge impact on the Gas

Turbine thermal efficiency and component life. Water washing frequency and good

filtration will extend the life of the turbine components. Enhanced air filtration will also

directly save fuel, improve machine power output and improve the reliability and

availability of engines.

Recognising that the gas turbine provides a unique challenge to the filter designer

commercial factors such as lifetime cycle costs along with operational resistance, filter

system efficiency and dust holding capacity must be considered with particular attention

to the volume of contaminated air consumed in a given period.

It is well established that conventional F8 / F9 filters satisfy the GT OEM aims of

acceptable hot end component life of at least 20,000 hours and also provides a power

output at a predicted heat rate and efficiency for a given inlet pressure loss over the

filtration system. With these parameters in mind the focus from the GT OEM is often to

keep the capital cost of the filtration system to a minimum to protect sales in a

competitive market.

3

efficiency from each new variant of gas turbine. The demand for increased performance

criteria and power output has generally resulted in higher firing temperatures and the

necessity for inter cooling of high pressure nozzles and other hot components. As a

result, the air quality has become even more influential with respect to machine

availability and life time performance.

The resistance of an air filter device or system has long been recognised to influence the

gas turbine power output and heat rate but what has generally been overlooked is the

impact of fouling on the compressor stages (Gas Generator).

It is generally accepted that high inlet resistance forces the gas generator to do more

work as it compensates for inadequate air flow. It is also recognised but difficult to

quantify, that the ingestion of sub 5 micron (μm) particulates, which impinge on the LP

compressor blades, cause a ‘fouling’ phenomena which in turn further deteriorates the

engine efficiency.

What has become apparent to some machine users who have increased the efficiency of

their air filter systems is that a higher system resistance has had little negative impact.

More importantly the consistent cleanliness of the compressor has reflected in a

significant improvement in all round performance of the gas turbine. In other words,

compressor fouling appears to be more influential in the health, life and economics of the

engine than inlet resistance.

Enhanced Hepa filtration undoubtedly reduces fouling and helps maintain compressor

efficiency but this does need to be balanced with the additional system pressure loss,

environmental conditions and the type of machine operation.

An introduction to the effects of poor air quality on a gas turbine is described below.

Modern gas turbine rotating parts are complex in design and structure and have a critical

profile for maximum working efficiency. The high pressure blades / nozzles sometimes

have small air holes to deliver cooling air as the working temperatures are close to the

limit of the material. Compressor blades are made of a very sophisticated alloy of metals

to provide strength and durability and these are coated with a protective layer for

durability. This makes effective filtration a major factor to the long term life of the gas

turbine.

Particles, which have sufficient mass to irreversibly wear the internal rotating

components, are typically identified as being greater than 10 μm in diameter. Their

manner. Such particles can be removed by inertial filters or pre-filters with

consummate ease.

Those pollutants which are less than 5 μm diameter do not have sufficient mass to

cause wear, but they can impinge onto the surface of the rotating and static components

and in a short time period change the blade profile away from its ideal shape. This is

4

temperature of components. Loss of compressor efficiency is normally compensated in

the short term with higher firing temperatures or increased compressor speed. This

phenomenon of fouling is reversible and is addressed by water-washing using

detergents and copious quantities of fresh water.

Photograph 1: The black deposits on the compressor blades is fouling caused by hydrocarbon

contaminants in the atmosphere

Corrosion of the LP and HP parts of the Gas Turbine is a risk if airborne salts pass

through the filter system. It is a chemical process which is not dependant on the

particulate size but on the presence of moisture and an electrolytic reaction between

salts and metals of different types. Airborne salt and water ingestion causes low

temperature corrosion whilst the combination of NaCl with air/fuel borne sulphur results

in high temperature sulphidation/oxidation or ‘hot gas’ corrosion.

Hot Gas Corrosion is of particular concern especially in coastal and offshore locations

where NaCl is prevalent both as a dry particle and in solution in water. When mixed with

sour (sulphurous) fuel it will cause accelerated degradation of key hot section

components.

To protect the rotating machinery from the impact of fouling, erosion or corrosion, gas

turbine manufacturers (OEM’s) issue mandatory air quality requirements to filtration

suppliers. The level of these requirements is not particularly stringent but also takes into

consideration that regular water wash and maintenance of the gas turbine will also be

required. For original equipment supply this enables the OEM to remain commercially

viable in a competitive market whilst balancing the performance, health and life of the

turbo-machinery.

5

In order to achieve combustion air cleanliness as specified by the machine Original

Equipment Manufacturer (OEM), gas turbines have traditionally employed barrier filters

which provide an efficiency level of F8 / F9 to European test standard EN 779:2002 (or

MERV15 / 16 to the American ASHRAE 52.2 test standard).

European filter classifications are covered by two standards EN779:2002 and EN1822

the classifications of which are summarised below:-

Coarse Dust Filter

G1 <65 (A)

G2 65-80 (A)

G3 80-90 (A)

G4 >90 (A)

Fine Dust Filter

F5 40-60 (E)

F6 60-80 (E)

F7 80-90 (E)

F8 90-95 (E)

F9 >95 (E)

High Efficiency

Particulate Air Filter

(HEPA)

H10 85

H11 95

H12 99.5

H13 99.95

H14 99.995

Ultra Low

Penetration Air Filter

(ULPA)

U15 99.9995

U16 99.99995

U17 99.999995

EN779 air filter test standard challenges the fine dust filter with a DEHS oil droplet

aerosol after multiple ASHRAE dust loading steps up to a given pressure drop while

coarse dust filters are tested with dry dust. In 2002 the standard introduced a discharged

efficiency to ensure a clearer filter performance was published, rather than an efficiency

which was still influenced by the electrostatic charge from a newly manufactured

synthetic filter.

For the High efficiency Particulate air filter EN1822 does not challenge any operational

life since there is no measurement on the dust loading capacity. EN1822 determines the

most penetrating particle size (MPPS), in clean condition only, and this is used as the

basis to determine filter classification H10 to U17.

6

are individually tested in a dry duct environment and real life operation and performance

There is currently no recognised international standard for testing filters in wet conditions

to quantify the filters resistance to water in a dynamic situation. This is a particularly

important factor and should not be underestimated for filters applied on gas turbine

intakes. A filter could have a good efficiency classification but if it was not impervious to

water, salt (in solution) could migrate through the filter. Over time when the

environmental conditions change the water would evaporate resulting in salt crystalline

growth downstream of the filters and ingestion by the gas turbine. In time this would lead

to compressor fouling, corrosion and if the air or fuel also has a high sulphur content, hot

gas corrosion with a consequential reduction in hot end component life.

Note: EN 779:2002 and EN1822 standards relate to air filters in air conditioning plant.

There is no existing recognised Filtration ISO, CEN or ASHRAE standard relative to

Turbo Machinery. However ISO is currently working towards developing a new standard

specifically for testing and classifying filters for use on turbo machinery. This is

scheduled for release by end 2010.

techniques of particle attraction and diffusion. A major component of this technique is the

air speed past the fibres and the diameter of those fibres. This means that a lower airstream

velocity will result in improved particle removal efficiency. Optimum filter media

areas are determined by test and it is recognised that pleat shape and size contribute

greatly to the overall performance of the filter.

Fundamental to filter selection is the recognition that all filter stages upstream of the final

filter are employed as pre-filters to maximize the final ‘fine’ filter life and suitable weather

protection is provided to limit the ingestion of rain, fog, ice and snow.

For static, non pulse, filter systems the HEPA class filter stage is most of the time an

additional 3rd stage. The HEPA stage (typically H10-H13) shall be protected with

“normal” up-stream stages typically 1st stage prefilter type G3-F6 and 2nd stage F8-F9.

Commercial and practical restrictions occasionally force alternatives to this and in some

instances, only one pre-filter stage can be selected.

Single stage reverse jet pulse filters are best suited to high dust laden environments but

currently no such product exists which will give a true Hepa efficiency on a single stage

self-cleaning, so they are to be considered as pre-filters to protect a final stage of Hepa

class filters.

Consequently provision of Hepa filtration protection to the engine normally requires an

additional stage of filtration over and above that employed to meet the GT OEM

mandatory requirements. This increases the inlet system resistance over the filtration

system however this can be alleviated by increasing the filter surface and may result in a

larger filter package. The total filter system capital investment will be higher than a

7

investment.

Experience already exists on land-based and offshore installations with Hepa grade

filters referenced as H10-H13 efficiency. These are much more efficient at sub-micron

particulate removal than the traditional F8 & F9 filtration systems. Of course special

treatments to prevent hydrocarbons interacting negatively with the filter and techniques

for rapid water removal from the inlet together with elimination of water penetration

through the fine filter is also essential. To help appreciate the step change that Hepa

filtration can offer, please refer to the attached comparison in Appendix A which

demonstrates how H12 vs F8 Filtration will result in the quality of the combustion air

ingested by the gas turbine being 1000 times (sub micron) cleaner.

To highlight the benefit of Hepa filtration, two examples of improved systems are detailed

below: 25MW turbine with H12 filtration and a 45MW turbine with H10 filtration.

Due to an improved level of filtration this example highlights the operational commercial

benefit of increased revenue through reduced downtime for offline water washing. The

analysis does not take into account the additional cost benefit associated with the life

extension of Hot end components and the consequential reduction in engine removal,

upgrade and off line refurbishment activities.

Gas Turbine Operational Cost Analysis – 25 MW Machine

Engine wash frequency – Hours 750 2000 8000

Expected filter life – months 24 24 12

Note 1 – The costs for washing and production are from a recognized North Sea Operator

Note 2 The example is to show the potential benefit to the operator of applying H12 filtration and

relates to a specific type of installation where the production is constant.

8

Original 2-stage filter system - F6 & F9

The 45 MW turbine was originally provided with a F6 pre-filter stage and a F9 final stage.

Target production for this application 45MW

During 22 weeks of operation the turbine was frequently on-lined washed >30 times with

no improvement of performance and two off-line wash @ 4 hours with only minor

improvement of performance.

Total loss of power during 22 weeks operation; 2300 MW.

Due to performance loss the economical impact was an income reduction of Euro

172,000.

Improved filter system - F7 & H10

The 45 MW turbine was provided with a F7 pre-filter stage and a H10 final stage.

22 weeks of operation with no off-line and no on-line washing.

Negligible power loss measured and target production of 45MW reached during the 22

weeks of operation.

Investment of improved filter efficiency <10% of Euro 172,000.

This example shows the need to analyse local conditions to best optimise the filter

system and that the filter system arrangement needs the possibility to be easily modified

after installation on site. This example also shows the need to balance pressure drop

and efficiency. For this site the higher final stage efficiency only decreased the power

output due to pressure drop, in addition with a slightly higher filter price.

Feedback from some operators that have added Hepa filtration, indicate that the

increase in filter resistance has not been problematic and the impact on engine heat rate

and power output has been minimal. The cleaner combustion air has prevented

deterioration in performance by avoiding compressor fouling and so the engine thermal

efficiency has remained stable. However depending upon the application other users are

focused to minimise the impact of pressure drop associated with an additional filter

stage.

Experience has proved that whilst Hepa filters will have a higher capital investment cost

and the inlet system has to be designed to achieve the best solution, the benefits are

huge in comparison. Extended hot end component life, improved availability and

increased revenues can potentially reduce filter pay-back in time to days.

9

levels used on rotating machinery with huge financial and technical benefits to the user

which greatly exceed the additional capital cost and cost of the consumable filters.

In summary, clean air can advantageously change the economics of Turbo-machinery

operation:

However, it must be considered that long ‘hot end’ component life has not necessarily

been in the interest of all parties in the supply chain of capital equipment and not all

sectors of the industry have recognised the benefits of high quality air filtration.

Technical solutions for high performing filter system has been available for decades, it’s

more a matter of willingness to invest from OEM’s and user’s side

It is also important to note that even higher air quality than H10-H13, which are the most

used Hepa alternatives, can be provided which is way in excess of anything required by

rotating machinery. European test standards, EN779:2002 & EN1822 together list 17

grades of filter efficiency from G1 through to U17. Many levels of even higher efficiency

filters can also be provided. The gas turbine industry is not stretching the capabilities in

air filtration technology but it can benefit by it without risk.

10

Note - The above curves are typical only and are provided to help give an appreciation of

the step change in efficiency moving from F8 to H12 classification.

Consider 1,000,000 particles size 0.5μm diameter upstream of the filter

F8 Initial efficiency at 0.5μm ~ 60%, therefore penetration = 400,000 particles

H12 Initial Efficiency at 0.5μm ~ 99.98%, therefore penetration = 200 particles

Comparison H12 vs F8; 400,000/200 = 2000 more efficient at 0.5μm

Consider 1,000,000 particles size 0.3μm diameter upstream of the filter

F8 Initial efficiency at 0.3μm ~ 50%, therefore penetration = 500,000 particles

H12 Initial Efficiency at 0.3μm ~ 99.95%, therefore penetration = 500 particles

Comparison H12 vs F8; 500,000/500 = 1000 more efficient at 0.3μm