Man this actually is a difficult problem. To pump just 5ml, a tiny 5ml! But you still have a 0.5m to 1m head to overcome and you need to do this several times an hour (as I recall). Hmmmm. Some random lateral thoughts:
- If the water were flowing, being pumped around a circuit and at the top you had a small 'take off' point and maybe a solenoid valve or motorised ball valve so you could drop a sample out of the flow. This way you could use an impeller pump which would last forever and the higher flow rate would not be an issue but maybe an advantage.
- Could you take this the other way around: move the sampling point into the tank? Move the whole analysis thing down to tank level so you don't need a pump at all.
- Use a container to bring the sample up rather than a pump? Maybe lower and raise a sampling container with some mechanical arrangement of miniature winch type thing or a bucket elevator actually might be the thing to do.
Apart from the risk of splashing I'm thinking the bucket elevator is a pretty damn fine option.
Otherwise all I can say is that I think you picked the right pump technology for the sample sizes at least so if you have to stick with a peristaltic pump you might have to invest time and research the tube technologies.
I didn't ask before; when the pumps died, what went wrong with them? Which bit failed? Tube or motor or gear box?
PE