One pilot recently told me this happened to him.
Somehow water got in and froze, still on the ground, and then the system obviously did not work when airborne.
The circumstances were unusual. The aircraft had wing covers on the wings, and it sounds like these ended up having water pooling inside them, and the water then forcing its way into the TKS panels, where it froze later on when the temperature dropped.
The panels contain some porous plastic which soaks up the fluid and prevents it all draining to the lowest point when the pump turns off, but there might be some cavities in there.
It would seem that it is a good idea to test the system periodically, to keep the panels filled with TKS fluid.
And maybe it doesn’t happen without wing covers?
I think a periodic test of the panels is essential in the winter to expel any moisture and in the summer to lift the remains of the bugs you have hit ( the gliding fraternity say bugs cost them speed and have in flight bug removal systems )
Peter wrote:
It would seem that it is a good idea to test the system periodically, to keep the panels filled with TKS fluid.
Checking fluid flow to the panels was part of the preflight for the decided TB20 I flew.
I have full FIKI on my SR22T. As part of the maintenance of the system you are required to peform a FIKI system preflight every 30 days winter or summer. This involved priming both pumps, running the system in normal mode (30 on 90 sec off) and then in backup mode (ie continuous). This will include inspecting that the porous panel membranes are ‘wetted’.
The reason per the SOP is that the panels may dry out and can cause uneven fluid flow during subsequent operation.
I presume by doing this you would also prevent the water contamination problem.
I time this with VOR checks that are required every 30 days on N reg IFR operations. If I am not flying into known ice I will do this in the air to save the mess on the hangers or taxi ways.
Sounds like the wing covers were a bit contra-productive in this case. All year round the TKS should be used every 2-3 weeks or so to keep it from clogging up.
I have only flown with TKS on a Cirrus and it works good.
Checking fluid flow to the panels was part of the preflight for the decided TB20 I flew.
What were the exact instructions for this (if any)?
I ask because a ground test leaves a big puddle of sticky fluid on the ground, even if you turn it off the instant you see the fluid coming out.
And if you run the system until the fluid is coming out of every portion of every panel, you could be dumping several litres on the ground.
I have just comprehensively tested the system after it was drained (for the replacement of the fluid level sensor) and it took about a gallon (say 4 litres) before it was coming out everywhere. I did that in flight, obviously, but it still took about 10 of these
full of hot water to get the stuff off the plane, after I landed.
All year round the TKS should be used every 2-3 weeks or so to keep it from clogging up.
I have to agree, but it leaves quite a mess, unless you are fortunate to have a hosepipe with hot water running through it 
Peter wrote:
What were the exact instructions for this (if any)?
I ask because a ground test leaves a big puddle of sticky fluid on the ground, even if you turn it off the instant you see the fluid coming out.
And if you run the system until the fluid is coming out of every portion of every panel, you could be dumping several litres on the ground.
I don’t recall that happening. Maybe it is different with the factory installed system?
Tks has always been and will always be messy. No way around that. It’s simply one of the major downsides. Whether one really needs to wash off the residue from the aircraft after each use is a much discussed topic.
Interesting PDF; many thanks. Funny they want to see all panels tested if you expect to have to use it
I think the amount of fluid wasted to test all panels depends on how long since you last used it. After some weeks the stuff does seem to drain out (largely due to wing dihedral) causing a lag when it is used again. So on planes with minimal dihedral (e.g. Cessnas) this may not be an issue.
I don’t think one must wash it; it is just that
On our Diamonds we cycle the system every Monday. There are times where the dust block elements and we have to put the system on MAX to ‘blow through’.. Indeed, even that didn’t work last week for one wing in a very smoggy environment. I’ve never experienced water ingress and associated ice.
I find that good housekeeping helps and we try to wash the leading edges every time we wash to windscreen, ie every day.
