I've been asked about surveying water well locations for well registration. A colleague sent me a copy of the letter submitted by a surveyor on a previous job. The surveyor performed static GPS observations and processed via OPUS, according to his letter report. In the stamped letter he makes the following statement regarding accuracy:
"the maximum Relative Positional Accuracy resulting from the survey does not exceed the Allowable Relative Positional Accuracy 0.07 feet + 50 ppm for horizontal location and Class 1 (0.01' +/-) for vertical location."
How would you interpret his vertical accuracy statement? My colleague asked if I could give him elevations to the hundredth. I said I could give you a value to the hundredth, but accuracy to 0.01' is not really possible, or perhaps not practical, with typical survey GPS systems. Even with a fairly long static session (4 hrs), this isn't possible, is it?? What should I tell him?
Elevation from where? I have been around a bunch of water wells and there isn't a really exact consistent place to measure the elevation that precisely. A lot of them they do a quick and dirty concrete slab around it with PVC sticking up above the concrete and then they saw it off once the concrete cures. Why on earth do the need a water well elevation that precise anyway? The driller logging it sure ain't working with those types of tolerances.
I drilled a second water well at my place for some irrigation last year and when we were permitting it the water district came out and "surveyed" it in with a Trimble Geo-XH GIS grade unit.
These are monitoring wells. For groundwater contamination sampling. You need an accurate elevation to determine groundwater flow direction. But 0.01' isn't realistic in my opinion.
What is important is elevation differences, not elevations. Simultaneous observations, specifying the same CORS on each for OPUS, will give better differences than sequential observations, and better differences than absolute elevations. Whether these will meet 0.01' I don't know.
How far apart are these wells? Far enough to think about ionospheric effects being different?
And is that 0.01 standard error, 0.01 at 95% confidence, or an absolute guarantee against any future more accurate measurement?
For this particular survey there were 14 wells on one site. Hard to tell exactly how far apart they were since his map has no scale, but less than a mile it looks like.
Bill those are my questions too about accuracy. I'm trying to find the regulation to see exactly how it's worded.
Ok here's the survey requirements:
6.1 Determination of Well Location
The vertical and horizontal position of each installation in the system shall be determined and subsequently mapped. The well location map should include the location of all monitoring wells in the system and their respective identification number, elevation of the top of the riser position to be used as the reference point for water level measurement and the elevation of the reference point on the slab. The locations and elevations of all permanent benchmarks and pertinent boundary markers located on-site or used in the survey should also be plotted on the map.
The water level measurement reference point should be permanently marked, for instance, by cutting a V- notch into the top edge of the riser pipe. This vertical reference point should be surveyed to the nearest one hundredth (0.01) of a foot in reference to the nearest NGVD reference point.
The horizontal location of all installations, active or decommissioned, should be determined by reference to a standardized grid to the nearest one tenth (0.1) of a foot. For LDOTD registration purposes, the latitude and longitude of the wells shall be determined to the nearest one (1) second of a degree.
Geographic Positioning System (GPS) surveys may be used if they meet the appropriate accuracy or are previously approved by the appropriate regulatory authority.
Ok so SURVEYED to the nearest 0.01', not absolutely accurate to that. Makes more sense.
andrewm, post: 363136, member: 10888 wrote: For this particular survey there were 14 wells on one site. Hard to tell exactly how far apart they were since his map has no scale, but less than a mile it looks like.
Bill those are my questions too about accuracy. I'm trying to find the regulation to see exactly how it's worded.
I've done peizometer locations on dams (monitoring wells for water seeping through the embankment) They wanted the elevation tight but where OK with the horizontal no so tight. So after a discussion with the project engineer we settled on RTK horizontal and a digital level loop through the peizomiters (top of pipe) that tied into benchmarks past the ends of the dams. The main benchmark was established by static GPS so the peizometer elevations where based upon that but all was relative. My digital level loops closed at a 0.01 feet so I was good calling it to that.
You need to talk to the end user of the data and find out what they really need and what it will cost to provide it. GPS RTK or Static WILL NOT give you 0.01 feet in elevation so you should tell them that you will need a good level loop for that and it will cost $$. I think on my project the digital level loop was 6 times the cost of the RTK horizontal. When you are done simply compare the GPS elevations to the level loop and you will have your accuracy answer.
andrewm, post: 363138, member: 10888 wrote: Ok here's the survey requirements:
6.1 Determination of Well Location
The vertical and horizontal position of each installation in the system shall be determined and subsequently mapped. The well location map should include the location of all monitoring wells in the system and their respective identification number, elevation of the top of the riser position to be used as the reference point for water level measurement and the elevation of the reference point on the slab. The locations and elevations of all permanent benchmarks and pertinent boundary markers located on-site or used in the survey should also be plotted on the map.
The water level measurement reference point should be permanently marked, for instance, by cutting a V- notch into the top edge of the riser pipe. This vertical reference point should be surveyed to the nearest one hundredth (0.01) of a foot in reference to the nearest NGVD reference point.
The horizontal location of all installations, active or decommissioned, should be determined by reference to a standardized grid to the nearest one tenth (0.1) of a foot. For LDOTD registration purposes, the latitude and longitude of the wells shall be determined to the nearest one (1) second of a degree.
Geographic Positioning System (GPS) surveys may be used if they meet the appropriate accuracy or are previously approved by the appropriate regulatory authority.
Ok so SURVEYED to the nearest 0.01', not absolutely accurate to that. Makes more sense.
We do these occasionally. We use RTK to establish horizontal, and level through the monuments, so that the relative hts between wells is very precise.
The tie to a known and published benchmark is done in various ways.
Andrew, I do a ton of these for several different firms. They all have tolerances for 1ft horizontal and .01 vertical. I use static gnss for 15 minute occupation with 2 bases on known points. I then check into a 3rd with the "rover". Do I get 0.01? Well we all agree that's silly but I get close.
You could level if you want. But I can do 15 to 20 a day with static and the effort level is low for 1 guy. I can get to about 0.03 very to be honest.
The Hydrogeoligists I work with are more interested in the relative elevation between wells than a datum tie to 0.01'. GPS horizontal and differential level loops for vertical.
DDSM:beer:
Thanks for the feedback. What do you think about the previous surveyor's vertical accuracy statement? Seems rather misleading to me. I would probably say my survey meets the blah blah requirements then regurgitate parts of the regs pertaining to accuracy. Or maybe I'm reading too much into his wording.
andrewm, post: 363138, member: 10888 wrote: This vertical reference point should be surveyed to the nearest one hundredth (0.01) of a foot in reference to the nearest NGVD reference point.
NGVD? That would be NGVD29!
Good luck with THAT (using GPS).
I'm not saying that it's not possible, BUT it does put another wrinkle into the equation.
To get anywhere near 0.01 feet relative to some remote USC&GS, USGS, NGS, [whatever] Bench Mark, is going to REQUIRE leveling, and possibly orthometric corrections.
Me thinks that the "spec" is silly, but if you sign off on it, then ya bought it.
Just my two bits
Loyal
I had to run elevations with a B-1 level for vertical from a bench mark. The N and S locations were Lat. and Long. Then I had to show the distance to two of the nearest section lines or property lines. I forget if they wanted other things.
Any guess as to how precise they are able to guage the depth to water in the monitoring well? Any stretch in that string?
You will need to level to meet the spec.
Explain this to the contracting entity and make sure they under stand GPS observations alone can't meet the spec (because somebody who believes their data collector will observe with RTK and report their ortho heights as meeting the elevation spec).
Water levels are measured with something like this:
http://m.solinst.com/products/level-measurement-devices/101/ds-101-p7.php
Very accurate. Pretty easy to get 0.01'.
Those look great! Have never witnessed anything used more precise than what amounts to a bobber on a heavy fishing cord.
I've done dozens of these projects. If you don't crack out the level (and use sound procedure) you won't meet spec.
andrewm, post: 363186, member: 10888 wrote: Water levels are measured with something like this:
http://m.solinst.com/products/level-measurement-devices/101/ds-101-p7.php
Very accurate. Pretty easy to get 0.01'.
We compare our water level meters with a Lufkin Highway Nubian every year and 'test' the operator's ability.
DDSM:beer:
