faking the water meter

My aerospace engineer brother is doing a little experiment. He lets the tap drip and collect the water in a pail, and see if this way it can save on meter reading.

Experimental setup wise, there is no basis of comparison that I'm aware of, so let's say it's flawed. And theoretically speaking I think you can fake the meter, but you need some research.

If I'm not wrong, the household water meter we have is of the rotary piston type. Water flows in from the inlet, turns the piston, and flows out in the outlet hole. As the piston rotates, some network of gears will turn the digit.

So, how to let water flow without turning the digit? You can't. But at either very high or very low flow rate, the meter becomes inaccurate. Of course you want it to undercount and not overcount, so what do you do? You get hold of the meter's characteristic chart and voila, the magical flow rate is 1.5 litre/hr.

Those who do maths would have come across something like "when this tends to zero, that tends to infinity" stuff. I hope in this case it's saving that tends to infinity and not cost.

(please don't play with the meter itself, they don't put a seal there for nothing. I know you are thinking of inverting the meter. Do you think the meter reading will go backwards?)