How accurate is a simple water level?

[couldabin]

I've used a 5-gallon bucket, filled with water, with two clear plastic hoses plumbed to the bottom to establish level over a distance of about 5 meters. It worked well enough for the circumstances.

Now I'm interested in establishing level over a distance that is closer to 30 meters, and would like to know if the two points are level within 3 cm or so.

Will this simple water level be that accurate? What are the factors that limit/enhance the accuracy? Does the diameter of the hoses affect it? What about the amount of water head (if, for instance, instead of a 5-gallon bucket I were to use a piece of large diameter PVC, 2 meters long)? What about the diameter at the end of each hose -- if the diameter of the last foot of the hose were made much larger, would it improve the accuracy?

TIA

 

[DaveC426913]

I've used a 5-gallon bucket, filled with water, with two clear plastic hoses plumbed to the bottom to establish level over a distance of about 5 meters. It worked well enough for the circumstances.

Huh. I hadn't thought of puttiong a bucket in the middle. when I leveled my garden, I used only a length of clear hose.

Now I'm interested in establishing level over a distance that is closer to 30 meters, and would like to know if the two points are level within 3 cm or so.

I can't imagine why it wouldn't be.

What about the diameter at the end of each hose -- if the diameter of the last foot of the hose were made much larger, would it improve the accuracy?TIA

Seems to me, making the diameter of the end smaller is what you want. That way, a small change is more noticeable, allowing finer measuring.

 

[FredGarvin]

It will be as accurate as you can be in your measurement. On it's own, the water is very hard to compress so over long distances and heights you won't see any real variations.

The diameter of the hoses simply dictate how much it weighs (within reason). The water will always find it's true level. The things to watch out for is to not have air in the line and to not have any pinches/kinks in the line.

 

[DaveC426913]

The one factor I can see wanting to be careful of is water viscosity/surface tension in relation to pipe diameter. I'm nor saying it would be a practical factor, just something that could come into play at the limits of experimentation.

 

[couldabin]

Dave, with regard to surface tension, would larger diameter mitigate the inaccuracy?

 

[DaveC426913]

Dave, with regard to surface tension, would larger diameter mitigate the inaccuracy?

I was thinking you might get some sort of wicking effect, but really I'm just talking through my hat.

Nonetheless, I would imagine you'd want it small but not too small. I think 1/2" would be minimum.

 

[berkeman]

I remember trying to use a store-bought water level some years ago (about 10m of clear 1cm tubing, as I remember, and it didn't work for diddly. When I held up the two ends next to each other, the viscosity and wetting at the ends of the water column made a significant error. Not as much as the 3cm the OP is asking about, but more than 1mm, which was way too much for the work I was doing. I ended up using long straightedges and careful level work to transfer the level line a moderate distance. Nowadays, I'd just use a good laser level.

 

[couldabin]

I recall reading somewhere that the relationship between the surface area of the primary water supply (my 5-gal bucket in this case) and that of the tube ends has a bearing on the friction/viscosity issues. The greater the difference the less the friction effect. At least, that's what I think I read. Does that make sense?

 

[chemisttree]

The ends of the hose should be wetted to the same degree provided that the material of the hose at the ends are identical. If one end is wetted (the end you put the water in) and the other end is dry, you could see some error but if both ends have been wetted and are clean, the accuracy is AMAZING... certainly on the order of a mm over that distance. You MUST be sure that the fluid is equally degassed throughout the length of the hose. This is equivalent to saying that the fluid has the same density from one end to the other. Water clouded by entrained air bubbles is slightly less dense than degassed water. Same goes for temperature. Bubbles that adhere to the hose and don't obstruct the flow will not cause an inaccuracy. Free floating bubbles will. Its best to remove them if possible. The use of two different sized hoses at the ends will have no effect on the accuracy. It is the height that is important, not the diameter since the pressure exerted on both ends is normalized to area.

...if the diameter of the last foot of the hose were made much larger, would it improve the accuracy?

It will improve the speed at which the level settles down for a reading.

 

[couldabin]

Thanks, all, for the input. I think I can put off going high-tech for now and rely on the old bucket.