Contents

Introduction

The ‘Kerosene Fungus’ has become of importance because of its association with aviation fuels. It was Hendey in 1964 who coined the name ‘kerosene fungus’ but prior to this the fungus had been known as the ‘creosote fungus’ (Marsden, 1954) because of its association with creosoted timbers. This fungus exists in two states; the imperfect or asexual state Cladosporium resinae (Lindau) de Vries (de Vries, 1955), and the perfect or sexual state Amorphotheca resinae Parbery (Parbery, 1969a). It is usually referred to as Cladosporium resinae because this is the state in which it normally occurs in kerosene and soil. To further complicate matters four forms of the imperfect state have been described. Intermediate forms exist and most workers find the fungus to be very variable indeed (see Parbery, 1969a).

Interest in this fungus was first aroused by reports of its occurrence in storage and aircraft fuel tanks containing aviation fuel in the early 1960's. Until the mid-1960's, however, a certain amount of secrecy or mystery surrounded the problem of microbial growth in fuels because much of the work was done at defence establishments or in technical laboratories and most reports were difficult of access.

Publicity could, perhaps, have resulted in public panic. The first significant publication in the open scientific literature which drew the attention of mycologists to the problem was Hendey's paper (he was working at the Admiralty Materials Laboratory, Poole, Dorset, England) published in 1964 which dealt with the distribution of the ‘kerosene fungus’ in kerosene-type fuels, its identification and the conditions under which growth occurs. The utilisation of hydrocarbons by micro-organisms had been known for some time previously and Bushnell and Haas (1941) showed that gasoline (= petrol), kerosene, light and heavy mineral oils and paraffin wax could be used as sources of carbon. They examined the ‘water-bottoms’ from various petroleum storage tanks but recorded only bacteria (Pseudomonas and Corynebacterium).

Lansdown, writing in the Royal Aeronautical Society's Journal in 1965, has listed specific problems associated with microbial growth in aviation gasoline and kerosene. These include fuel pump failures and corrosion, filter clogging and fuel tank corrosion. Both bacteria and fungi were implicated. He said at that time (1965), ‘It has now become apparent that microbial contamination is widespread in aircraft fuel supply systems, both on land and in aircraft carriers, where serious clogging of fuel system filters has occurred.’ The problem, although worse in the tropics, appeared to be world wide. It has now been shown that among the fungi the ‘kerosene fungus’ is the organism most frequently implicated, e.g. it was present in 78% of all fuel samples from aircraft tanks examinated in Australia (Hazzard, 1963) and in 80% of all fuel samples examined in California (Engel and Swatek, 1966). It is probably the most important micro-organism in contamination of fuels and in corrosion at the present time (Parbery, 1968).

In addition to its presence in aviation fuel the ‘kerosene fungus’ has been found in other hydrocarbon fuels and hydraulic fluids (Nicot and Zakartchenko, 1966), on creosoted timbers such as telegraph poles and railway sleepers (Marsden, 1954), on tarred woods, asphalt and resins (Christensen, Kaufert, Schmitz and Allison, 1942), bituminised cardboard (Anon, 1968), a cosmetic face cream (de Vries, 1952), the female sex hormone, progesterone (Fonken, Murray and Reineke, 1960) and methyl-p-hydroxybenzoate (Solkolski, Chidester and Honeywell, 1965).

Very recently it has been isolated from soil in Australia and Europe (Parbery, 1968a, 1969b) and in New Zealand (Sheridan, Steel and Knox, 1971) and it appears to be widespread.

Both Hazzard and Hendey have indicated that a complete study of this fungus is an immense task (Hazzard, 1963). Our knowledge can be expected to increase only gradually as more mycologists become interested in the fungus and initiate studies on its biology. In our opinion the three most significant contributions of recent research to our knowledge of this fungus are: (1) the development page 23 of a reliable method for isolation of the fungus from the soil (Parbery, 1967), (2) the finding of the perfect state in Australia (Parbery, 1969a) and New Zealand (Sheridan and Steel, 1971), and (3) the development of a reliable method for isolating the fungus from the air (Sheridan and Nelson, 1971). The finding of the perfect state is particularly significant because the sexual state allows development of new strains of the fungus. Now that a reliable method is available for isolating the fungus from the air it is possible to monitor air and determine when and how much fungus is present. Earlier workers were of the opinion that the fungus is airborne (Christensen et al., 1942; Hendey, 1964) and it possibly enters the fuel via the air, but they were unable to isolate it from the air. In our laboratory we are currently studying isolates from fuels, soils and air, a number of which produce the perfect state under mineral oil.

We became interested in the ‘kerosene fungus’ as a result of finding it in a local soil in November, 1969, Our studies have been concerned with the biology of New Zealand isolates of the fungus, their distribution in soils and occurrence in air and aviation fuels. The two questions which are put to us most frequently are (1) how widespread is contamination of aviation fuels with C. resinae at the present time and (2) how does the fungus get into the fuels? These are obviously important questions. The former is difficult to answer because of inaccessibility of reports as mentioned earlier and the expression of conflicting opinions, the latter because of the failure of earlier workers to isolate the fungus from soil or air. These questions, and others, have prompted us to make a thorough search of the literature in an attempt to gather together in one paper a body of information on the ‘kerosene fungus’ and its activities. In so far as our own work here in New Zealand contributes to these aims we refer to it. The paper is published in three parts. Part I is concerned with the problem of microbial contamination of aviation fuels, Part II with the natural habitat of C. resinae and Part III with morphology and physiology of C. resinae.

1.Background to the Problem

For many years now micro-organisms have been known to grow in hydrocrabons (Miyoshi, 1895; Kaserer, 1906; Stormer, 1908; Sohngen, 1913), but it is only relatively recently that their growth has caused concern. It is the consequences of their growth that raises problems, not just their presence. Fuel pumps and filters may be clogged, capacitance gauges may not function properly and tanks and pumps may be corroded resulting in engine failure (Rogers and Kaplan, 1967).

The problem of microbial contamination of jet fuels was first observed in the American Air Force during 1956 (Leathen and page 24 Kinsel, 1963). Results of studies by Bakanauskas (1958) indicated that micro-organisms were present in the sludge from the fuel tanks. Analyses of sludges from aircraft based at a Tropical Air Force base showed that sludges contained in addition to micro-organisms, microbiological debris and metallic products, break-down products of the hydrocarbons, iron oxides, surfactants, chlorides, metallic salts, water, silica and other extraneous materials (Churchill, 1963). Many of these contaminants had probably entered the fuels in handling between the refinery and the consumer — foreign matter such as soil, sand, dust and dirt, metal chips, cellulose fibres and metal oxides (Rogers and Kaplan, 1967). Others were the result of corrosion within the fuel system.

Parbery (1968) has noted that except for one report made in 1944 most reports of the microbial corrosion of aluminium have been published since 1960. These deal almost exclusively with the corrosion of alloys in fuel tanks and wing components of aircraft using jet fuel. Integral fuel tank corrosion was first observed in the American Air Force in 1960. About the same time similar problems were encountered in several other military and commercial airlines and in the Australian Air Force and Australian airlines (Churchill, 1963). However, it was not until 1960 and 1961 when the American Air Force began its modification programme on the relatively new ‘wet wing’ aircraft (B-52 G, KC-135 and C-130) that the seriousness of integral fuel tank contamination and associated corrosion problems was recognised (Ward, 1963). Large amounts of sludge and bacterial growth were found in the tank bottoms and sides — corrosion was widespread in the areas covered by this contamination. The problem was greatest in tropical environments. Ward (1963) reports that the C-130 fleet has had aircraft with holes corroded completely through the lower skin from integral fuel tank corrosion. Corrosion ranged from isolated pits and small areas of exfoliation to extensive pitting and exfoliation covering complete panels. The wing tank skin was made of aluminium plates and stringers of extruded, integrally stiffened aluminium alloy panels. No corrosion-resistant coatings were provided for the tanks at that time. The shape of the tanks made it virtually impossible to drain off water and sludge.

In addition to the above there are a number of other reports of problems associated with microbial contamination of aircraft fuels. These have been gathered together by Lansdown (1965) and include: fuel pump failures in Valetta and Hastings aircraft operating from the Suez Canal Zone in 1952; heavy fuel pump corrosion in Canberra aircraft operating from Butterworth in Malaya in 1956; fuel filter clogging in Boeing B-47 and KC-97 aircraft in 1958 and fuel tank corrosion in North American Fury (FJ-3 and FJ-4) in 1961.

Hazzard (1961) has reported fuel tank corrosion in Lockheed Hercules and Electras in Australia. The problem has also been reported from Brazil (Gutheil, 1966).

From these reports it is clear that damage has occurred to Australian, American and British aircraft. Most trouble is encountered where aircraft are operating in tropical regions and in aircraft with integral fuel tanks. Damage to British-built and British-operated aircraft, according to Lansdown (1965) has been extremely low and in all cases where it occurred aircraft had been operating in the Far East. The damage was disclosed during maintenance without any operating difficulties arising. He suggests as a possible reason the temperate climate of Britain which gives a lower daily temperature variation and thus a lower water condensation (water is essential for growth of micro-organisms in fuels) and temperatures below the optimum for microbial growth. Also the tank drains, to remove water and sludge, appear properly located in British aircraft. There is another possibility worth considering, that is the possibility that different strains of micro-organisms may be involved. This aspect has not yet been studied as far as we are aware although we have found some evidence for the existence of strains of the ‘kerosene fungus’ differing in their growth rate in kerosene (see Part III).

2.Method of Attack By Micro-Organisms

Aviation fuel consists almost entirely of hydrocarbons (compounds containing only carbon and hydrogen) but will also contain traces of additives and contaminants. In high-octane aviation gasoline the major additive is a lead compound, usually lead-tetraethyl with an organic bromide to prevent lead fouling. Other additives present in much smaller quantities in aviation fuels are anti-oxidants, metal de-activators and possibly corrosion inhibitors, anti-icing additives or anti-static additives (see Lansdown, 1965). Modern jet aircraft use aviation turbine kerosene or kerosene-type fuels. In its natural state kerosene can dissolve up to 75 parts per million (ppm) of water, the actual amount depending on temperature and atmospheric humidity. Free water will be present in fuels subject to temperature variations due to condensation. This water will extract components from the fuel and might, for example, contain a few ppm of hydrocarbons and several per cent of anti-icing additive. Both hydrocarbon and water phases will contain dissolved air (Hill, 1970) and materials leached from tank coatings or produced by corrosion of the metals present.

Sea water has often been deliberately introduced to fuel tanks to seal off leaks at the bottom of the tank, to flush out the fuel from less accessible parts of the system and to act as ballast in tanks. For example, it has recently been claimed that at least 20% of the world's tanker fleet discharges 600,000 tons of oil during tanker cleaning — the equivalent of five Torrey Canyons (Anon., 1970a). The sea water in the tanks will provide inorganic salts and possibly biological contaminants.

Even in the absence of sea water fuels may contain fungal spores and bacteria presumably derived from the air or soil as contaminants during handling. These are not removed by normal filtration (but see under ‘Methods of Control’). Some of these organisms will arrive at the interface between fuel and water: the fuel provides the source of energy, the water provides dissolved oxygen and trace materials. In vented aircraft and fuel tanks air is drawn in to replace fuel used and airborne fungi and bacteria can thus enter and may settle in the fuel. In gasometer-type storage tanks (with a floating roof) rain water carrying microbial contaminants may enter the tanks (Prince, 1961).

Different classes of hydrocarbons are preferentially attacked by different micro-organisms. Hendey (1964) has shown, for example, that the ‘kerosene fungus’ Cladosporium resinae can use kerosene as its sole carbon source. This has been confirmed in France (Nicot and Zakartchenko, 1966) and New Zealand (Sheridan, Steel and Knox, 1971). Between 20% and 50% of the carbon assimilated by bacteria and fungi is converted into cell substance (Zobell, 1946), the proportion being higher in the earlier stages of growth.

The remainder of the carbon is converted to more highly oxidised compounds including carbon dioxide, organic acids, alcohols and esters. According to Lansdown (1965) these products will themselves alter the environment: e.g. lower molecular weight fatty acids will lower the pH of the aqueous phase making it more corrosive to metals. Reactions with specific mineral salts may produce more corrosive acids. Corrosion of metals gives rise to more available ions in solution (De Gray and Killian, 1960). Higher molecular weight fatty acids are surface active and by concentrating at the fuel/water interface will reduce the interfacial tension and permit ready emulsi-fication of fuel and water. Alcohols and esters will increase the solubility of fuel in the aqueous phase. All this results in extension of the zone for optimum microbial growth. Other micro-organisms which hitherto had remained dormant may now find conditions suitable for their growth. It may be that growth is now rapid and oxygen is all consumed leading to anaerobic conditions. If sulphate is present (from direct contamination or oxidation of sulphur-compounds in fuel by aerobic micro-organisms) conditions will be favourable for the growth of sulphate-reducing bacteria. These produce fuel-soluble corrosive sulphide (e.g. hydrogen sulphide) which can be carried with the fuel and cause corrosion of components of an aircraft fuel system (Hill, 1970).

Although it has been claimed that saline water, rust, sulphates and ‘dirt’ will themselves cause corrosion even in the absence of micro-organisms (Calvelli, 1963), a strong microbial growth will most probably enhance corrosion (Lansdown, 1965). London, Finefrock page 27 and Killian (1964) have questioned the correlation of microorganisms with malfunction but since their work was published there is ample evidence to show a definite correlation.

Corrosion of aircraft integral fuel tanks has been studied by a number of workers (see Parbery, 1968) and without doubt has been very serious. In some cases wing planks have been corroded through (Lansdown, 1965). Where tanks were lined with rubber coatings it appeared that fungi were able to penetrate these and there is also the possibility that they could do the same to the metal walls of the tanks. Lansdown (1965) points out that although there have been indications that mechanical penetration of aluminium foil can occur, it is very difficult to show that the effect is not a simple chemical attack associated with the corrosive products of metabolism around the growing fungal hyphae. Experiments by Hazzard (1963) have shown that the fungus Cladosporium resinae can penetrate rubber linings (see details under ‘Methods of Control’). It is well known that many plant pathogens can mechanically penetrate thin films of chemically inert materials by the formation of special attachment organs (appressoria) (Wood, 1967). C. resinae has been observed by us to attach itself firmly to the glass walls of culture bottles: the attachment organs closely resemble appressoria, so purely mechanical penetration would not be surprising. No work has yet been done on this aspect in our laboratory.

Microbial attack is also shown by the formation of sludge or solid matter which may clog downstream parts of the fuel system, particularly filters and screens. There is some doubt as to whether bacterial slime has sufficient mechanical strength to block filters (Bakanauskas, 1958) but there is little doubt that fungal mycelium can block filters, screens and other small apertures, perhaps even the drain points of fuel tanks. Pump screens partly blocked by fungal growth have been found by Hazzard (1963).

Fungal growth may also become attached to the fuel tank walls and prove difficult to dislodge during cleaning operations.

De Gray and Killian (1960) report that some rust inhibitors appear to function as nutrients for bacteria. They have found stimulation of growth at normal levels of such inhibitors (25 to 50 ppm) in both gasoline and kerosene. In a kerosene, essentially the same as a JP-4 fuel, they found the bacterial count after 48 hours without rust inhibitor to be 50,000/ml.; with 25 ppm inhibitor it was 2,200,000/ml. The surface-active rust inhibitors apparently reduce the interfacial tension of water and hydrocarbon, and so increase the availability of hydrocarbon to bacteria. ‘Bacteria encourage rust and make slimes, which hold rust in suspension, which encourages bacteria, which encourage rust and make slime …’ is a very apt description of the vicious circle (De Gray and Killian, 1960). Anti-icing additives may in low concentrations also prove nutritional to micro-organisms (Hill, 1970).

3. The Relative Importance of Bacteria and Fungi in Fuel Contamination and Corrosion

The following organisms are reported as having been isolated from jet fuel and jet fuel-water samples:

Among the fungi Cladosporium was the most prevalent and among the bacteria Pseudomonas aeruginosa.

Parbery (1968b) has reviewed the literature comparing the roles of bacteria and fungi in corroding aluminium and concludes that bacteria were generally believed to be more important although he considers that there is no acceptable basis for this belief. Numerous difficulties are encountered in attempting to equate growth of bacteria and fungi because the former are unicellular and readily counted, the latter are mycelial (thread-like) and tend to clump. Reports of much of the work done in an attempt to elucidate the problem are not available to other research workers because they have not been published in the open scientific literature, a situation which may, in effect, mean that ‘they are to all intents and purposes not published’ (Parbery, 1968c). It is clear, however, that masses of bacterial growth are more easily broken up than fungal growth and so are less likely to cause blockages. Fungal hyphae can readily attach them-selves to solid objects such as fuel tank walls and may exert a mechanical pressure which may allow penetration of the linings (see under ‘3. Method of attack by micro-organisms). Bacteria cannot do this. The microbial growth sludge removed from corroded fuel tanks invariably contains a mixture of fungal and bacterial species (Parbery, 1968b). The bacterium Ps. acruginosa and the fungus C. resinae have been found often in association with corroded alloys and each can induce corrosion in aluminium plates in culture (Hedrick, Carrol, Owen and Pritchard, 1963; Davis, 1967).

Parbery (1968b) has studied growth of C. resinae in relation to corrosion of aluminium and found a correlation between amount of mycelium produced and the percentage loss in weight of aluminium pieces. Hendey (1964) showed that all samples of aluminium in his experiments which were in contact with C. resinae were perforated and had lost considerable weight: controls were unaffected and had lost no weight.

Other organisms less commonly found, which also apparently cause corrosion, are a gram-ve bacterium identified as ‘17-11’ (Hosteller and Powers, 1963), Desulfovibrio desulfuricans (Hedrick, Carroll, Owen and Pritchard, 1963; Iverson, 1967) and A. niger (Hedrick, Carroll, Owen and Pritchard, 1963).

Many organisms, other than the above, found in fuel may possibly be dormant and do not grow — some of the fungi listed at the beginning of this section possibly arise as contaminants during sampling.

Hazzard (1963) has surveyed 45 airfields in Australia and the Far East and examined 599 samples of aviation turbine kerosene. He found C. resinae in 218, Paecilomyces in 55, A. niger in 48 and Helminthosporium in 27.

Rogers and Kaplan (1964) studied the growth of over 800 isolates of fungi from fuels in relation to their ability to grow in the presence of aviation fuel. Best growth occurred with isolates of C. resinae, Fusarium sp. Stysanus sp. (mis-spelled as Stypanus) and page 30 Alternaria sp. A group classed as having moderate ability to utilise fuel included several Cephalosporium spp.; other Cladosporium spp. and a few penicillia, aspergilli and an Helminthosporium. Basidiomy-cetes, A. niger, Trichoderma sp., Pullularia sp., yeasts and actinomy-cetes grew very little or not at all. Nicot and Zakartchenko (1966) have inoculated sterile minimal medium with a mixture of fungi isolated from jet fuel and covered the medium with a layer of fuel. Only C. resinae grew.

Figure 1. These bottles contain an aqueous mineral salts medium with a layer of aviation turbine kerosene on top. The fungal mat on the end of the needle has been removed from the right-hand bottle and represents 10 weeks’ growth of Cladosporium resinae. The left-hand bottle is an uninoculated control.

Engel and Swatek (1966) report that Ustilago (a smut fungus) was isolated from a number of fuel samples (but not from soil). They report that all their isolates of Ustilago grew readily with fuel as their carbon source, sporulating profusely. Ustilago is a plant pathogen and its ability to grow in aviation fuel appears rather unusual.

We have examined a number of fungi isolated from aviation fuels in New Zealand for ability to grow in creosote and in aviation kerosene (unpublished). These were identified by us as Cephalo-sporium sp., Paecilomyces sp., Penicillium spp., Trichoderma viride, Stemphylium sp. and Cladosporium resinae. All except Stemphylium and T. viride grew on an agar medium containing 0.1% creosote; only C. resinae grew on the medium containing 1% creosote and it was the only fungus to grow in aviation kerosene. (Fig. 1.)

Isolates of C. resinae from soil and air also grew readily in kerosene. None of the yeasts isolated by us and only two of the bacteria (belonging to the genus Pseudomonas) grew in kerosene. We are currently studying corrosion.

It seems clear, therefore, that the fungus C. resinae is undoubtedly the most important and dangerous contaminant of aviation fuels at the present time. However, as Hedrick, Reynolds and Crum (1968) point out, there is an urgent need for intensive studies on fungi (moulds and yeasts) and bacteria in connection with their growth in jet fuels because our knowledge is still very small.

4. The Occurrence of C. Resinae in Aviation Fuels

Figures for the occurrence of C. resinae in aviation fuel in aircraft and storage tanks are not readily available. However, a summary of those available to us is given here.

Engel and Swatek (1966) studied the micro-organisms isolated from 59 jet aircraft integral fuel tanks in California, U.S.A., and found the most prevalent fungus to be C. resinae (= Hormodendrum resinae). It was present in 80% of the samples. They mention that of 9 soil samples taken from strategic positions at the overhaul base, none yielded this organism. This is rather surprising since the fungus has subsequently been shown to be widespread in soils in Australia (Parbery, 1969b) and New Zealand (Sheridan, Steel and Knox, 1971), and would be expected to enter fuel tanks from this source via the air.

In an examination of fuel tanks of Australian aircraft during 1961 and 1962 Hazzard (1963) found fungi in all but one (a relatively new Boeing 707) of 21 aircraft. C. resinae was the predominant fungus isolated: it was present in 78% of all samples. His results for 157 samples from tanks of the 21 aircraft are shown below.

He reports that Shell examined 65 samples of fuel from 2 Comet IV, 6 Boeing 707 and 4 Electra aircraft and found C. resinae in 63 of them (97%). Contamination was heavy with counts of 100-300/litre (see ‘6. Methods for monitoring fuel’).

In a later survey of fuels (1963) delivered to civil aircraft in Australia and Pacific Islands, carried out by private companies, page 32 C. resinae was reported in only 21 samples out of 194. This represents 39 airfields. The counts were less than 10/litre. The survey was carried out after attempts had been made to control the problem.

Because of the high incidence and levels of contamination originally found in military fuel systems and aircraft, the R.A.A.F. arranged overhaul and where necessary replacement of all fuel installations at its bases. Subsequent testing (in 1963) showed a very much lower level of contamination: C. resinae was found in only 9 out of 88 samples taken from 10 bases. Previous figures were 47 out of 115. Fungal counts were now down to 3/litre.

These results indicate a considerable reduction in observations of the occurrence of C. resinae consequent on the application of measures designed to reduce the problem.

Reliable figures for the occurrence of C. resinae in aviation fuels are not yet available for New Zealand. Some indication of the present situation in this country may, perhaps, be reflected in our recent examination of 13 samples of aviation turbine kerosene, 6 of which were found to be contaminated with C. resinae. The highest count was 24/litre.

The catalogue of the culture collection of the Commonwealth Mycological Institute, Kew, (Anon., 1968) lists C. resinae from kerosene filters from Brazil, Japan, Syria, Nigeria, Denmark, England and India. It is also listed from aircraft fuel tanks from New Zealand.

5. Methods of Control

Present methods of control are aimed mainly at preventing or destroying growth of micro-organisms in the fuel. This can be tackled in a number of ways: (1) regular removal of water, which is essential for growth, and periodic washing out of tanks to remove microbial growth; (2) addition of a biostat which prevents reproduction, or biocide which kills micro-organisms; (3) use of resistant tank coatings so that the metal walls of the tanks cannot be reached by the micro-organisms or their metabolic products; (4) filtration to remove all bacterial and fungal growth and (5) removal from fuels of fractions essential for microbial growth.

The first approach constitutes what is commonly known as ‘good housekeeping’ and is the usual method of control, as far as aviation kerosene is concerned, applied at present. Where applied rigorously it works satisfactorily. But difficulties arise, for example in aircraft carriers, or in oil tankers, where fuel is stored for long periods. It is impossible to eliminate water entirely from aircraft fuel tanks because condensation provides a constant source of free water. The best that can be done is to remove this water by frequent and careful draining. Tanks are now designed to make this possible. Filters and filter separators have been designed to remove solid contaminants and water from the fuel supply. In order to kill page 33 micro-organisms in some cases tanks may be washed out with 2% potassium dichromate. According to Lansdown (1965) these ‘good housekeeping’ measures have virtually eliminated microbial corrosion from Australian and American aircraft. They do not appear, however, to have eliminated the fungus C. resinae (see 4. The occurrence of C. resinae in aviation fuels). Their value lies in constant and careful application.

The second approach, i.e. the addition of biostats which prevent reproduction or biocides which kill micro-organisms, is not so easy to apply. The biostat or biocide must be active chiefly in the aqueous phase yet must not be removed from the system by the periodic drain checks which are essential to remove condensed water. In a modern jet aircraft such as the Boeing 727 it is estimated that between one pint and one quart of water is produced for each vent in the fuel tank every time an aircraft lands (personal communication — Shell (N.Z.)). Hence best control is achieved by using water-soluble/non fuel-soluble biocides in fuel storage installations. Compounds such as silver wool, borates, chlorides and methyl violet (Lansdown, 1965) have been tried for water bottoms where sulphate reducing bacteria occur, and a boron compound, Biobor JF, has been used in some United States executive aircraft.

An anti-icing inhibitor containing 99.6% methyl cellosolve and 0.4% glycerol (= ethylene glycol monomethyl ether or EGME) introduced by the U.S.A.F. in 1962 appears to act as a biocide (Lansdown, 1965; Hill, 1970). Neither methyl cellosolve nor glycerol are active on their own. Hill (1970) has shown EGME to be an effective biocide but concentration of EGME and degree of contamination of fuels by micro-organisms is important. He found that at 10% concentration in mineral salts medium EGME required 7 days to effect sterilisation. On the other hand an organo-borate at 1% produced sterility within 24 hours. He found that in low concentrations EGME may even be actively nutritional. He further describes experiments in which mixtures of para hydroxy-benzoates dissolved in EGME are used as a fuel biocide, and sterility achieved within a few hours. Another drawback to the use of inhibitors is their high cost (Hill, personal communication).

The only figures available to us are as follows. The price of Biobor, marketed in the United States by Boron and Chemical Corp., is given as 60 cents/lb. in bulk, $1.21/lb. in case lots (Anon., 1970b). It has been estimated that initial treatment at 270 ppm would cost $2.19/1,000 gal. fuel. Maintenance levels of 135 ppm would run at $1.10/1,000 gal.

Rogers and Kaplan (1967) have screened a number of prospective biocides, tabulated the requirements for the ideal biocide and discussed two commonly in use which meet most of the criteria. These are EGME and a mixture of boron compounds.

From the beginning of 1970 all the R.A.F. fuels have contained EGME (Hill, personal communication). The Australian services and the R.N.Z.A.F. use an anti-icing agent which contains EGME and has apparently eliminated the problem of microbial contamination in their aircraft (personal communications).

In the third approach, the use of resistant fuel tank linings, much work has been done. According to Ward (1963) a polyurethane coating has successfully withstood many tests devised to simulate microbiological environments within fuel tanks. Hazzard (1963) compared a number of tank lining materials, recommended by airframe manufacturers: these included Buna-N rubber, polyurethane and epoxy top-coatings. Fungal penetration (by C. resinae) of Buna-N occurred within 4 to 6 days but no penetration of epoxy or polyurethane had occurred after 148 or 180 days respectively. The incorporation of fungicides into Buna-N films delays penetration. He recommended the use of polyurethane in integral tank protection. On the other hand, Miller, Herron, Krigrens, Cameron and Terry (1964) found that a polyurethane film was penetrated in 8 days by a mixed culture of micro-organisms obtained from a JP-4 fuel tank. They suggest that the polyester type of polyurethane may be less resistant than the polyether type. Hazzard was working in Australia, Miller et al. in America: the discrepancy of results may also be possibly due to the existence of species or strains of fungus differing in their penetrating ability. This aspect has not been studied by anyone as far as we are aware. A ‘furane’ lining has been claimed by McGregor (1963) to give protection.

In the fourth approach, i.e. filtration to remove all bacterial and fungal growth, Hedrick, Owen, Carroll, Pritchard, Albrecht and Martel (1964) have concluded from their studies that fuels treated with ultrafine mechanical filtration can be made free of microbial contamination. They used 0.35 and 0.45 micron pore-diameter filters. Approximately 13.8 square feet of filter surface was required to filter 600 gallons/minute of JP-4 fuel at 50 p.s.i. filter differential pressure. About 20 square feet would be required to filter JP-5 under the same conditions. These figures are for fuel with no free water — the presence of water slows down the rate of filtration considerably and may affect sterility. They concluded that sterility could only be obtained providing all water is removed from the fuel prior to filtration. Fuel additives may affect the efficiency of the filter membrane. Normal filling (not using an ultrafine filter) of a VC-10, for example, is at the rate of 1,000 gals./min. Even after removal of all micro-organisms from fuel, contamination may subsequently occur from air sucked into tanks through vents during flight. Recent work (Sheridan and Nelson, 1971) has shown that C. resinae, the most serious fungal contaminant, is air-borne.

Before the fifth line of attack can be applied (removal of fractions of fuel required for growth of micro-organisms) it is necessary to page 35 determine what these fractions are. Prince (1961) has examined aviation kerosene before and after growth of the fungus C. resinae using gas chromatography. It appears that this fungus utilises many components of the kerosene and it appears unlikely, therefore, that removal of a specific fraction from the fuel would have much effect. Further work is necessary.

Other methods of control which have been tried include gamma ray sterilisation and radio frequency (Kivel, Chatterton and Palchak 1967; Hedrick, Carroll and Owen, 1964).

In conclusion, therefore, it can be said that ‘good housekeeping’ applied rigorously has proved an effective method of controlling microbial contamination of aviation fuels and consequent problems. It seems unlikely, however, that this method will succeed in eliminating the fungus C. resinae from fuels because of the widespread occurrence of the fungus in soils and in the atmosphere. Indeed, now that strains of the fungus, present naturally in soils in New Zealand, have been shown to grow more vigorously than many of those isolated from fuels (see Part III) a constant watch must be kept for any increase in the incidence and levels of C. resinae in aviation kerosene.

The problem of microbiological fuel contamination and corrosion is now generally considered to be under control. Continuing care is necessary, however, to maintain this control. The search for effective biocides still continues although to date there has not been much success. A combination of ‘good housekeeping’ with good aircraft design ensuring well-located drain points, tank linings which are resistant to fungal penetration, and an endeavour to keep fuel as clean and dry as possible is still absolutely essential if trouble is to be avoided in the future.

7. Summary and Conclusions

There have been a number of reports concerning problems raised by the growth of micro-organisms in aviation fuels. These include filter clogging and corrosion of pumps and fuel tanks. If left unchecked, there is little doubt that the consequences of microbial growth in fuels would be very serious. Both bacteria and fungi have been isolated and some have been shown to use the hydrocarbons as a food and to cause corrosion. Among the former Pseudomonas aeruginosa is the most prevalent, and among the latter the ‘kerosene fungus’, Cladosporium resinae. The incidence and level of C. resinae in fuel tanks and systems was high in the early 1960's but have fallen considerably in recent years as a result of the vigorous application of control measures, particularly ‘good housekeeping’. Few serious problems occur at present. Nevertheless there is no reason for complacency. Some people appear to believe that because the biological population is concentrated in the aqueous phase, the elimination of water or the suppression of life in the water will eliminate the micro-organisms. Hazzard (1967) has found that this view is incorrect. Some organisms, notably C. resinae, can introduce into the hydrocarbon phase material which will remain viable for several years in essentially water-free petroleum products. The application of ‘good housekeeping’ procedures will not, therefore, eliminate the micro-organisms from fuel. The demonstration that many soil and air isolates of this fungus grow more vigorously in kerosene than those normally isolated from this source further illustrates the seriousness of the problem (see Part III).

There is an urgent need for research aimed at developing better methods of detecting microbial growth in fuels and better methods of controlling it. An intensive study of all those organisms capable of utilising hydrocarbons is needed. The work of Parbery in Australia and our own work here in New Zealand is directed towards obtaining a better understanding of the biology of the ‘kerosene fungus’, C. resinae. Parts II and III of this paper are concerned with this work.

Acknowledgments

The authors are grateful to the Chief Superintendent of the Defence Standards Laboratories, Australia, for permission to quote data contained in D.S.L. Report No. 252, and to the Chief Chemist, B.P. (N.Z.) Ltd., for supplying samples of aviation kerosene. Thanks are due to Mr. Herbert Christophers for photographic assistance and to Mrs. Mary Sheridan for reading and criticising the manuscript.