Home Documents Images Message Board
(Use your browser's back button to return to the page that you were previously viewing.)


May 1983
Safety Board says flaws allow undrainable water in fuel

Reprinted from Aviation Safety Magazine, Vol 3, No. 5. Reprinted by permission.

Visit Aviation Safety Magazine's web-site at the following URL:

Copyright 1983, Belvoir Publications Inc. All rights reserved. Reproduction in whole or part is strictly prohibited.


Safety Board says flaws allow undrainable water in fuel


One article of faith for every careful airman has always been that if he conducts his preflight inspection religiously, particularly by [observing] the age-old rite of sumping the fuel tanks for evidence of contamination of water, he should not have a power interruption.

The National Transportation Safety Board has recently issued a sweeping series of recommendations that may shake that faith.

If taken to heart by the FAA, the recommendations would result in airworthiness action involving tens of thousands of aircraft.

In a nutshell, NTSB has asserted that with certain aircraft, even if the pilot performs a careful preflight, and even if he were to go to extreme additional measures, he still could not remove all the water that may have seeped into his fuel tanks.

The aircraft cited by NTSB are mainly Cessna single-engine models with bladder type fuel tanks, including the Cessna 180, 182, 185, 188, 206 and 207.

In other recommendations issued at the same time, NTSB also called for mandatory quick-drain retrofits to another series of airplanes--the Piper high-wings, such as the J-3 and PA-18 Cubs, the Super Cruiser and the Tri-Pacer. Also affected would be the PA-25 Pawnee airplane.

The recommendations were issued March 8 and at presstime, FAA had not prepared a response. FAA in January did issue a proposal to adopt an Airworthiness Directive on the Cessna 182 series, but not specifically and directly addressing the concerns advanced by NTSB.

At least a part of the impetus for the recommendations can be traced to information published in Aviation Safety nearly a year ago (June 1982, page 15). Reader-engineer Rodney Gross, who had experienced a water induced engine failure on takeoff in his Cessna 183 in 1978, related his discovery that water remains in the plane's fuel cells despite herculean efforts to remove it. He reported having developed a modification to sidetrack the water before it can get to the engine. Gross's STC for the modification was specifically mentioned by the NTSB in its recommendations to the FAA on March 8.

A Cessna spokesman at presstime said he could not comment, since the company had not yet received a copy of NTSB's recommendations.

Extent of Problem

In its recommendation letter, NTSB cited 396 cases of engine failure or malfunction, which occurred during the period 1975 through 1981 and involved water in the fuel as a cause or factor. It said these accidents primarily involved smaller, single-engine airplanes, most commonly on takeoff. The accidents resulted in the deaths of 72 persons, serious injuries to 93 persons, and minor injuries to 127 others. NTSB said the accidents "frequently involved older, high-wing Piper airplanes with metal fuel tanks…and high-wing Cessna airplanes, both old and new, with rubberized bladder-type fuel cells…" such as the aircraft mentioned above.

Upon an Aviation Safety request for raw data, NTSB provided a printout showing that of the 398 accidents cited, the Big Three aircraft manufactures had the following involvement: Cessna 155, Piper 87, Beech 14.

NTSB conceded that comparative rates of such accidents for each model of aircraft were not computed, partly due to inadequacies of data provide by FAA on flight-hours in recent years.

However, another rough measure of relative accident rates can be constructed by using the numbers of each aircraft in the FAA general aviation registry.

Using the NTSB listing as a staring point, Aviation Safety has computed a rough rate table for the various models of aircraft. The results are shown on page 3 [of the original publication. The results are contained in the table immediately following this paragraph in this reprint]. The rates can only be considered an approximation, since they are based on accidents per 1,000 registered aircraft, and do not account for the number of hours each aircraft is typically flown.


Accidents Involving Water in Fuel



Fleet Size

Rate per 1000 A/C









Cessna 150, 152




Cessna 170, 172, 175




Cessna 180, 185




Cessna 186




Cessna 182 (all)




Cessna 182P




Cessna 206, 207




Cessna 210




Other Cessna singles




Cessna twins




All Cessnas












Piper J-3




Piper PA-18




Piper PA-11, -12, -18




Piper PA-20, -22




Piper PA-25




Piper PA-28 series




Piper PA-32 series




Other Piper singles




Piper twins




All Pipers












Beech 19, 23, 24




Beech 33, 35, 36




Beech twins




All Beeches












All models








All GA aircraft






Accident data from NTSB printout of accidents involving water in fuel as a cause or factor, 1975-81 (1981 incomplete). Fleet size data from FAA Census of U.S. Civil Aircraft. December 31, 1981.



However, the comparison points clearly to certain findings:

  • The high-wing Pipers and the bladder-equipped Cessna singles do have a significantly greater rate of water-in-fuel accidents compared with the rest of the general aviation fleet.
  • Of the Pipers, the greatest problem appears to be with the PA-25 Pawnees.
  • Of the bladder-equipped Cessna's, the model 182 stands out as having a greater rate than other models, and a special listing shows specifically the model 182P to have the highest rate in the GA fleet.
  • There is hardly any involvement of twin-engine aircraft in water-in-fuel accidents.

Trapped Water

Water arrives in a fuel system in any of three ways. It may be pumped into the tanks with the gas by an FBO--a possibility that may be rare, but too common to ignore. It may leak through faulty caps when the plane is parked in the rain. It may condense from humid air.

All three ways are dangerous, since even if the condensation water only amounts to an ounce, this is sufficient to cause engine stoppage if it all goes to the engine at once. "Bad gas" from an FBO or rainwater leakage can put many ounces--even gallons--into an airplane's fuel system.

Because water is heavier than gasoline, it is assumed that a properly designed fuel system will employ gravity to allow the water to travel to the lowest point in a given tank or the total system, and there the designer will put a sump drain or quick-drain. The pilot arriving at the aircraft need only drain out this water and can be assured that what is left in the system is fuel.

NTSB's recommendations find this wrong.

[The beginning of this paragraph is not available.]…

…Pacer. Also that year, Piper issued a "Service 5parts Letter" (SP-6) which announced that wing fuel tank quick-drains were available for previously made Tri-Pacers, and for all PA-11, PA-12, PA-18, PA-18A and PA-20 series aircraft which had a quarter-inch tapered pipe fitting on the fuel tank.

Although a lot of the affected owners surely bought the quick-drain kits, the letter was never made mandatory, so there theoretically could be many older Tri-Pacers, Super Cubs and the like without wing quick-drains.

NTSB proposes that FAA issue an Airworthiness Directive ordering these drains to be installed.

As for the J-3 and for a similar situation in the PA-25 Pawnee airplane, there was no kit offered to retrofit a quick-drain on the aft lower portion of the fuselage tank. Nonetheless, NTSB proposes that FAA issue an AD ordering that such drains be installed.

NTSB has nothing critical to say about most later Pipers. "Modern Piper tricycle-gear airplanes, such as the PA-25, PA-32 and PA-38, which routinely incorporate fuel tank quick-drains have relatively few accidents involving water in the fuel compared to the older Piper airplanes." NTSB said.

[The beginning of this paragraph is not available.]…

…Injuries (a mild concussion, cuts and bruises).

When Gross got out of the hospital four days later, he went by to see the airplane and found mechanics still had not yet found out what had caused the engine failure.

The mechanics had concentrated on the engine, which was found to be capable of operation. Now, they took a look at the fuel system, and found virtually everything forward of the firewall (lines, gascolator, carburetor bowl) filled with water. Because the loss of the main gear at impact had ruptured the fuel lines under the cabin, the rest of the fuel system could not be checked.

However, the findings left Gross saddled with the signs of "pilot error." He knew other aviators would make the assumption that he had failed to perform his preflight correctly.

"But this didn't fit with the facts." Gross told us. In his preflight, "I knew I had done what pilots are supposed to do, and that didn't do the job."

Gross started "digging into it," and rather quickly found that contrary to what a pilot looking at the diagram in the plane's handbook might surmise, the gascolator area is not the fuel system low point. Gross analyzed the system and found the fuel lines, where they come down the aft doorposts of the cabin and run forward represent the approximate system low point. It was his hypothesis that water could be trapped in the U-shape areas formed by these lines.

Practical Test

Gross decided to test the theory. He obtained use of a Cessna 182, put an extra drain valve at this approximate low point area, leveled the plane and put the selector on the right tank, which was half-full of fuel. He now introduced quantities (in increments totaling about 76 ounces, or more than half a gallon) of water into the right tank, then tried to get the same amount of water out.

It was readily apparent that typical preflighting methods would not get the water out. Some pilots, going on the premise that "It all connects to here anyway," use only the Skylane's engine compartment gascolator-strainer drain, an unsafe practice. In imitation of this, Gross drained roughly a gallon and a half of fuel out of the strainer drain, but got not a trace of water.

Taking a sample from the newly added low-point drain did now show a trace of water (half an ounce in a pint's worth of fuel), but it was not the definitive outpouring of water he had hoped for. Gross would eventually discard this simple solution, in view of what he now found.

Now going to the wing drain, Gross collected lots of the water on the first try (47 ounces). But then the fuel ran clear for a further half a gallon of sampling.

Gross had now accomplished rather more draining than any typical pilot does on a preflight, and yet there was still more than a pint of water in the tank.

He now tried rocking the plane vigorously to encourage the water to flow to the drain--if it was going to. The water didn't. The first sample after rocking contained about three ounces of water in a pint of fuel, but then three more pints of fuel were drained--with rocking in between--and no water came out. Gross quit for the night with 15 ounces of water remaining in the tank.

The next day, nearly two quarts of clear fuel flowed from the wing drain without evidence of that water.

Gross siphoned off the top of the tank's fuel (but not the water), until there was about five gallons remaining in the tank. He now used the low-point drain to take out the five gallons. None of the water came out with it.

Finally, by jacking up the plane's right main wheel an inch and vigorously rocking the wings, Gross persuaded the residual fuel and the 15 ounces of water out of the tank.

Anyone would forgive a pilot for being too tired to fly if this were the kind of preflight he had to conduct!

It was the engineer's opinion (later seconded by NTSB) that slight ripples or ridges in the bottom of the Skylane's bladder tanks can act like small dams holding significant quantities of water back away from the tank drain as the plane sits at rest. However, inertial effects, such as a rolling takeoff, a slip or a skid, could roll the water over the dam and allow it to get to the fuel line.

Gross in 1979 initiated a lawsuit against Cessna seeking to recover the damages done by what he believes is a faulty fuel system design. He also set about to develop a solution to the problem.

The low-point drains alone, he had demonstrated, would not do the job. Gross eventually worked out an arrangement of [a] small reservoir, or "header" tanks placed in the belly of the airplane at the approximate low points of the system, with quick drains under them. Now, though water may collect in the wing tanks and be undrainable, if it later comes out, it will fall into the reservoir tanks and settle, allowing gasoline to pass above it. Gross eventually gained an STC for the system modification on all Skylane model variations in June of 1981.

Not Unique

Meanwhile, at several points around the country, others were having the same [suspicions] about the high-wing Cessnas and their bladder tanks. One lawyer representing a pilot involved in a crash reportedly put together a fuel system mock-up showing essentially what Gross had independently found and made a videotape record of it. In the spring of 1982, this lawyer informed NTSB of his findings.

NTSB officials contacted Cessna and asked the company to analyze the problem and suggest a solution.

At roughly the same time, NTSB investigators in the Atlanta field office were called to study a crash of a Cessna 206 that seemed to display the same problem.

News of Gross's STC and his experiments were relayed to the investigators by Aviation Safety.

Amid this context, Cessna now issued a rather unusual service letter to owners of the entire single-engine fleet (bladder-equipped or not). Owner Advisory SEB2-36A, issued July 30 1982, reminded owners to do proper preflight checks for water and other contaminants in the fuel. This was hardly newsworthy.

Tucked into the letter, however, was the advice that the pilot should "gently move the wings and/or lower the tail to the ground (on nose gear aircraft) to move the contaminants to the sampling points and assure that they are drained from the fuel system." (Our [Aviation Safety's] italics.)

Ineffective Answer

This is rather an extreme preflight procedure (It contained warnings about not harming the wings or tail while performing it), but if it worked, it would at least be an answer.

The Atlanta NTSB investigators decided to use Cessna's new procedure in a test of their own.

They put 32 ounces of water in a nearly full wing tank of a Cessna 182P; allowed it to settle, and checked the wing drain--no water evident.

They now rocked the wings and held the tail to the ground in accordance with the Cessna letter. Draining several quarts of fuel produced only five ounces of the water. They now held the tail down and jacked the left wing up, and in this manner got out another 10-12 ounces of water. But nothing would make the remaining 15-17 ounces of water come out of the tank. Eventually, they had to drain the entire tank and swab out the cell to get the remaining water out.

There was nothing apparently wrong with the tank, but its floor did contain small ridges and wrinkles.

NTSB apparently considered the investigators' test a little informal, so it arranged for a further test, conducted by the University of Illinois Institute of Aviation in February of this year, [1983]. The academics improved on the test by coloring the water red (the fuel was blue 100LL) and by measuring very precisely in milliliters. When all was said and done (including holding the tail down and jacking up the wing on a Cessna 182Q), some 500 ml (about 17 ounces) of water remained in its tank--trapped behind a prominent diagonal ridge on the floor of the cell, as well as in slight ridges and wrinkles elsewhere.

Clearly, Cessna's advice to owners will not solve the problem, NTSB said. "The Safety Board believes that the only reliable means of eliminating all of the water that may be entrapped within these fuel cells is to drain them completely, purge them with fresh air, and swab them to eliminate all traces of fuel and/or water, " NTSB said.

It added, "Other fuel system components, including fuel header tanks, carburetor bowls, fuel strainers and fuel lines between the strainer and the low point of the system, also should be drained at the same time."

NTSB called on the FAA to issue an AD to this effect, applying to bladder-equipped Cessnas from the Model 180 through the 207. Notably, it did not mention Cessna 210s of the strutted, pre -1967 vintage, which also had bladder tanks (this may be an NTSB oversight). All the models are similar in having bladder tanks and a very slight dihedral angle (about 1¼ to 1¼ degrees).

NTSB also suggested that a "fuel system design change is warranted," and it suggested that FAA look into possible solutions, specifically noting the STC held by Rodney Gross.

It should be mentioned that lots of other airplanes in the country have bladder tanks of almost exactly the same construction as the targeted Cessnas. However, these invariably have much larger dihedral angles--in the range of 3-5 degrees or greater. The larger Beech Bonanza and Baron fleets represent perhaps the greatest example. Perhaps because of the dihedral and other elements of the fuel system design, these airplanes do not seem to have the same problem at all as demonstrated by their low involvement in water-in-fuel accidents.

Focus on 182P

The involvement of the Cessna 182P model is remarkable and worth special attention. The 25 accidents involving 182P models constitute 64 percent of all the Skylane accidents (39) even though the 182P constitutes only 19 percent of the Skylane fleet.

The 182 started production in 1958 and went through various model changes (182A, 182B, etc.) over the years. The 182P was built in model years 1972 through 1976. Even though these latter-day Skylanes may fly more often than the other variants, they amount to only 2,605 aircraft of a total registered Skylane fleet of 13,682.

One NTSB investigator told us he couldn't account for this vastly greater rate of involvement, except by noting that the 182P heralded the advent of the optional long-range tank. Presumably, the longer the tank, the more water can be trapped by its wrinkles.

But Aviation Safety can point to another strongly possible factor: Prior to 1973, the largest manufacturer of bladder cells for Cessna --Goodyear--had been producing its BTC-39 model of fuel cell. Between 1973 and 1978, Goodyear produced a different model the BTC-67. The significant difference was the chemistry of the polyurethane rubber used in the cell. The BTC-67 was much stiffer. Its production period almost exactly matches that of the Cessna 182P. (After the P model, Cessna dropped bladders in Skylanes and the current 182Q model has a wet-wing tankage system.)

When a fuel cell is shipped, it is folded. The installer (whether at the factory or in the field) is given instructions to warm the cell in hot water to make it more pliable. The cell must be crumpled and stuffed into the wing opening. Its top is now secured beneath the top skin of the wing using snap fasteners.

The installer smoothes out [the] bottom of the cell as best he can, but only gravity holds it flat. It can easily be imagined that any wrinkles formed during storage, or while the cell was crumpled for insertion, will be left in the floor of the cell. If the cell material is stiffer, the wrinkles will be more prominent and less likely to smooth out over the years.

It should also be noted that Cessna has had its fill of bladder tanks. Starting around 1978, it has been converting its entire bladder-equipped fleet into "wet-wing" designs.

Further Aspects

While studying the bladder-tank, Cessnas, NTSB found other aspects of the fuel system that it believes also ought to be corrected. The primary element of concern is the Cessna fuel cap. This has a special history all its own.

Historically, many Cessna single-engine designs have incorporated a fuel tank in each wing with a venting system that links them together. This results in just one vent tube with which to allow air into the top of both tanks as fuel feeds out the bottom.

If the vent becomes obstructed, several ill effects may occur. The most serious and evident is a stoppage of fuel flow when the engine fuel pump can no longer draw. Prior to that, however, the engine can pump hard enough to create a vacuum in the tank that "sucks up" the floor of the bladder--which typically is not held down by anything except gravity. The now-[misshapen] tank may cause the fuel sensor to "think" there is more fuel present than there actually is. The pilot may not only be approaching fuel starvation when the tank no longer can feed, but could be approaching fuel exhaustion, with a gauge that indicates plenty of fuel in the tank.

To cope with this problem, Cessna in 1978 issued a service letter offering owners fuel caps that had vents incorporated in them. If the tank venting system failed, the cap vents would take over.

At about the same time, Cessna changed from its traditional metal fuel caps to a cap made of red plastic.

FAA in 1979 made the Cessna service letter into an AD (79-10-14), and tens of thousands of Cessna owners wound up buying vented plastic caps. Some owners, mistrustful of the plastic, were able to get vents put in their metal caps through either of two STC's that arose at the time.

The plastic caps have not done well in service, NTSB found. It checked FAA files of Service Difficulty Reports and discovered that particularly with the flush-style caps (found in the Cessna 180 and larger singles), the plastic is subject to warping, cracking and other problems.

NTSB quoted some of the reports: "warped red plastic fuel caps, sun exposure may be a factor": "plastic caps age and become brittle": "considerable amount of water found in fuel, apparently leaking through plastic caps during rains, seals appear good": "plastic fuel tank cap will not hold shape and will not seal with a new seal, fuel tank siphoned empty": "plastic crystallized from weather and broke."

NTSB also said the number of fuel cap SDRs increased after the plastic caps came into service in large numbers.

The consequences of leaky fuel caps could include both allowing rainwater into the tank, and allowing fuel to siphon out due to suction of airflow over the wing.

The Safety Board recommended that FAA conduct "an engineering evaluation of Cessna's flush plastic fuel caps to determine their sealing/venting characteristics under various critical service conditions, including extremes of temperature. If deficiencies are noted, appropriate corrective action should be required."

It appears that Cessna has not been unaware of the problem. In 1980, it issued a service letter outlining various steps to insure that the flush-type caps seal properly. The letter (SE80-59) stressed that the old metal caps have an O-ring seal (with a round cross-section) while the new plastic caps have a shaped gasket (nearly square in cross-section): Cessna cautioned that the two are not inter-changeable.

Cessna devoted renewed attention to the problem last year. In a service letter on July 23, 1983 (SE82-34) it advised of a procedure to check the fuel caps for proper sealing. (The mechanic plugs up the vent system and blows through a hose to pressurize the fuel tanks, then checks for leakage around the caps.)

FAA Proposal

Nor has FAA been entirely inattentive to the problems.

FAA in January of this year, [1983], did advance a proposal to issue an AD that would address part of the problem, for part of the Cessna Fleet. It suggested that an AD be adopted applying to Cessna 182 and 182RC models equipped with bladder tanks and a single primary vent system. It would call for a placard to be placed near the fuel gauges warning the pilot of the potential for erroneous readings. It would also call for repetitive inspections to cure problems of fuel loss, erroneous fuel quantity indications and undrainable water in the bladder cells."

However, NTSB said the FAA proposal would not go far enough. First the AD ought to apply to the entire range of bladder-equipped Cessnas, not just the 182s. Second, NTSB called for the fuel cap leakage test (per SE-82-34) to be conducted at the same time as the affected fleet's tanks are drained, purged and swabbed, and then periodically thereafter.


It is clear from the sweeping nature of the NTSB recommendations and the preliminary moves by the FAA that airworthiness action can be anticipated throughout the Cessna bladder-tanked fleet (and perhaps the Piper high-wings as well).

In the interim, the basic message for owners is that any suspicion of water in the fuel is grounds for extreme caution. A total fuel system purge is not out of the question for the owner who desires some peace of mind. Likewise, leaking fuel caps cannot be tolerated, since they will allow water to get in, and gas to get out.

Further, although the NTSB did not dwell on the subject, owners of all kinds of airplanes must consider other parts of the fuel system whenever water is detected in the tanks. Small droplets of water may have made their way to the carburetor bowl: they won't have any effect until they reach the top. In addition, Aviation Safety has noted reports of fuel injection controllers collecting small amounts of water, which can corrode the housing and eventually clog the controller.

And obviously, pilots who have become accustomed to rather cursory preflight sump checks will have a compelling reason to change their habits


May 1983
Safety Board says flaws allow undrainable water in fuel

Reprinted from Aviation Safety Magazine, Vol 3, No. 5. Reprinted by permission.

Visit Aviation Safety Magazine's web-site at the following URL:

Copyright 1983, Belvoir Publications Inc. All rights reserved. Reproduction in whole or part is strictly prohibited.

Home Documents Images Message Board
(Use your browser's back button to return to the page that you were previously viewing.)