Measuring water flow rates

[Chocolate fireguard]

The second reference given by langcliffe in his 12.01 post today mentioned the Constant Injection Method as being easy to understand but difficult to do in the field.
If I were wanting to do a measurement on a typical smallish fast-flowing mountain stream I think it`s the one I would start with. Because you have to start somewhere.
If you introduce salt at a known rate into a stream then at  points downstream the salt passes at the same rate (a bit like the electric current in a circuit being the same at all points I suppose). Obviously you have to allow time for this to happen, you have to sample far enough downstream for complete mixing to have occurred etc - no doubt these and other things will need thinking about.
But for a small fast turbulent stream with no big pools to delay things it should be possible to do a measurement when the introducing and sampling points are 100 metres or less apart, possibly taking much less than half an hour.
The sampling would need specialist equipment (to eventually come up with a concentration), but introduction could be done (accurately enough for trial purposes) with a big bag of salt, a teaspoon measure and a watch.

So if you introduce at 50g per minute and your titration/conductivity measurement says 10g/cubic metre you have 5 cubic metres of water per minute.
As always, repeatability will be a good guide to accuracy.

Feel free to scoff.

 

[langcliffe]

Feel free to scoff.

On the contrary, Chocolate fireguard, it's a technique which has served me well in the past. In the olden days we had to use titrametric analysis, but it should be much easier with these new fangled electrical conductivity meters. They don't seem that expensive, either.

 

[Fulk]

I wouldn't scoff, Chocolate fireguard, because I've used your method 'for real' to estimate the flow of a small river.

We made a big bucket of very concentrated salt solution, and tipped it in the stream at a known rate (something like a litre a minute) and sampled the flow downstream where the water had been thrououghly mixed. How did we know that it was thoroughly mixed? Becasue we added fluorescein to the concentrated solution, and estimated by eye the point at which the water was homogeneously green, and sampled downstream from there. Then it was a matter of titrating the concetnrated solution (after dilution) and the river samples with EDTA for Ca2+ concentration. The dilution factor taken with the 'flow rate' of the introduced salt solution gave the flow rate.

Problems? Well, the biggest problem was that some cows came along to see what we were doing, and one of them got her head in the bucket of salt ? she evidently liked salt, as she started to slurp it up, and we were treated to the spectacle of this cow, with bright green liquid dripping from its mouth.

A week or two later in the pub I obverheard some talk about green milk. 

 

[robjones]

When I was with Welsh Water Scientific Services in central Wales 25 to 30 years ago, fluorescein and rhodamine were widely used in dilution gauging to measure flow volumes of small rivers and streams where it was not necessary to install wooden vee-notch or half vee-notch weirs as part of medium-term monitoring.

Inevitably very visible mistakes could occur. The most memorable was when streams entering Frongoch Lead Mine were being assessed for a pollution alleviation sceme. Increasing amounts of fluorescein were put into the input streams but there was resolutely no trace in the resurgence from the mine. Ultimately, in desperation, all available  fluorescein was tipped in. The enormous volume of the flooded mine workings had held all the coloured input water and it all seemed to pop out in one go. The River Ystwyth turned green for some miles. Fortunately the main slug of green river water passed through the town of Abersytwyth at night and was dispersed at sea by morning.  :-[

 

[TheBitterEnd]

FWIW, Palmer (Cave Geology) has a section on flow measurement with a description of techniques on pages 102/103. He lists Volumetric, Float, Pitot tube Current Meter, Weir (V Notch) Manning Equation and Critical Flow over a flat topped weir. His view is that a V Notch is easiest (once installed) and one of the most accurate, he does not list any specific requirement for channel shape upstream.

 

[grahams]

I don't want to be the wielder of a large spanner but the OP's idea of measuring the amount of water going into sinks by the flow rate would appear to be impractical for his purposes. Flow rates for all the water dripping into minor sinks along the grit/limestone boundary and for water falling onto bare limestone could not be measured by this means.
A potential solution would be to measure the flow rate of a major stream sink in the area of interest and calculate the stream's catchment area to give an expected flow rate per km2. This would give a 'control' flow rate per km2 for the area. All you would then need to do would be to measure the flow rate at the resurgence and compare that with the total expected flow rate for the whole proposed catchment for the resurgence.
If this could be done with sufficient accuracy, the results could give information about the positions of underground watersheds and whether there are any as yet undiscovered Lady Blues (undetected resurgences) in our major rivers.

 

[darklord]

Grahams, I like that idea

I'm certainly not intending to test every drippy little shakehole on the contact, just a few of the obvious major feeders which would probably make up the bulk of the capacity (I do acknowledge that a still considerable amount might make its way into inlets via seepage and the sponge effect)

I'm really looking not so much for great accuracy but more a ready reckoner approach - so like, 100% is volume out of rising, now test a few main sinks....so approx. 30% from this one, 20% from that one etc. APPROXIMATELY....and I mean give or take 10% or so.

Such measurements, if they can be easily made, would be fine for my purposes. I'm not writing a PhD