On July 11, 1999, a Mooney M20F took off on its third short flight of the day from St. Andrews, MB with some members of a local Girl Guide troop as part of a larger event at a
Girl Guide Jamboree. As the aircraft passed through 340 feet AGL it was seen by witnesses on the ground to level off, roll to the left and enter a spin. The resulting crash fatally injured the pilot and the two passengers.
The investigation conducted by the Transportation Safety Board found the aircraft's engine likely stopped producing power due to water contamination. After the power loss, the pilot was unable to maintain sufficient airspeed to prevent a stall and the aircraft entered a spin from which the pilot didn't recover.
Why don't water and fuel mix?
Fuel and water are immiscible, that is to say, they won't mix. The reason the fluids don't mix has to do with their molecular structure; water is a polar solvent and fuel is non-polar solvent (read more about this here). Due to the difference in structure, water and fuel will never blend together; instead, when shaken or mixed, the water will be suspended in the fuel for a period of time after which, because the water is more dense than fuel, it will settle to the bottom of the container.
Checking for contamination
Checking for water contamination prior to every flight is a critical action. As part of the pre-flight inspection, the fuel system must be checked in accordance with the manufacturers instructions. This usually involves taking a small sample of fuel from each of the aircraft's gascolators. Small amounts of water will appear as small, clear bubbles on the bottom of the fuel sample while larger amounts will be clearly separated with water on the bottom and fuel on top. If water is found, you should continue to take samples until all the water is gone. Looking at the fuel sample against a constructing background can help you see contaminates. There have been cases where pilots have taken a fuel sample and assumed it was ok because it didn't have bubbles or a distinct separation but in reality, the entire sample was water. If you're ever unsure if you've got all the water out, check with another pilot or a maintenance engineer to be sure. It's not worth the risk of finding out in the air when the engine quits.
Once the sample is taken, you have to dispose of it. Some pour the sample back into the tank, not an issue if it's not contaminated and there are sample bottles contain a water filter that allow you to pour the fuel back into the tank without taking the risk of pouring water back in the tank; some pilots pour the sample out on the ground, I don't recommend this and in some cases it's illegal; and some pilots pour it into a safe container they keep in the hangar. Whichever method you choose the important thing is you check the tanks for water prior to flight. Be extra cautious when checking for water after the aircraft has been washed or left out in the rain or snow.
There are steps that can be taken to try and prevent water from entering the tanks in the first place, ensure the seals around the fuel caps are in good condition and are sealing properly and fill the tanks after flight; tanks left half empty or with space at the top allow water to condense on the exposed surfaces inside the tank then run into the fuel. It's also a good idea to buy fuel from reputable sources.
The other reason to sump the tanks...
Corrosion in the fuel system happens when water is left to collect in the lowest parts of the system. In most aircraft the lowest part of the system is where the gascolator is installed but some aircraft such as the Mooney in the case above have had issues with water collecting against ribs inside the fuel tank and, in that specific case, an Airworthiness Directive was issued to ensure drain holes that would allow trapped water to drain was issued.
Water trapped against a metal surface will start an electrochemical attack on the exposed metal surface. The corrosion will result the in the weakening of the metal structure if it's not stopped. This corrosion can occur inside the tank or within any of the lines or fittings between the tanks and the engine. The corrosion can introduce debris into the fuel system that can block lines, filters, and injectors that could, in the worst cases, cut off the fuel flow to the engine and cause it to quit.
Interestingly, in aircraft powered by jet fuel, the presence of trapped water can result in micro-biological corrosion where organisms can start to grow in the trapped water and feed off the jet fuel. If not stopped, these organisms will destroy the paint or surface protection of the metal tank and cause it to corrode. I remember a story of a Beach 1900 I used to fly that went in for routine maintenance. When the maintainers inspected the integral fuel tank they found a rag in the back of the tank that had become saturated with water over time. The resulting corrosion had worked its way through the tank wall and into the wing spar. The damage was so significant the aircraft was unable to be brought back into service and was written off.
Fuel system corrosion can be prevented (or at least minimized) by routinely draining the gascolators for each tank and ensuring there is no build-up of water.
Other safety information
Loss of control - In the accident case above, after the engine stopped producing power, the pilot allowed the speed to decay. Remember, job one is to fly the airplane, nothing else matters if you don't do this. Before troubleshooting or trying to communicate with ATC, make sure the aircraft is flying. After an engine failure, pitch and trim to maintain the best glide speed for your airplane, find a place to land and get setup on the approach then work on communication and trouble shooting.
Airworthiness directives (AD)- ensure that all the airworthiness directives for your aircraft are completed. Transport Canada has a search engine on their site that allows you to search for ADs by aircraft registration, while the results are not exhaustive, they are a great place to start. As part of your inspection cycle, go through the old ADs, figure out which ones have reoccurring actions, what action needs to be taken, and search for new ADs. It's not just a regulatory requirement, it's just plain smart. The pilot in the accident case above could have done everything possible to ensure there was no water in the fuel tanks prior to taking-off but because that airworthiness directive wasn't complete, it didn't matter.