Elevations - RTK vs Levelled

Neat experiment.  Thanks for posting.  I think these type of real world examples are VERY handy for people to see.

Had a conversation just last week with a friend who had marked some grade stakes for earthwork on a subdivision he is doing the surveying on (pretty much wide open field).  The contractor was complaining that the stakes weren't right because there was an inch ± of slop in them when he checked with his level.  My friend had used RTK to set the grades.  He was venting to me about it being 'close enough', but I had to side with the contractor - should have broken out the TS or level.

A little real world experiment.

I set out nails at ±300' intervals along a half mile of road and levelled through them, up and back, using a Leica DNA10 digital level. Loop closure was flat - zero.

Tied all the nails using a Leica GS18 base/rover pair. Employed the tilt sensing feature on the rover, so none of the shots had a plumb pole.  GPS base was between a half mile and several hundred feet of the tied points. Base location had good, but not great, sky. I held 2 of the levelled points for elevation.   

The road where the nails are is mostly tree lined and some of the locations were very poor for RTK. None of them are really what you might call good. Each RTK result is an average of at least two 3 second occupations. Weighted by the dc for coordinate quality. 

-- attachment is not available --
-- attachment is not available --

So you see that the elevations are fairly tight where the conditions approach good, and fall off where conditions are really bad. 

Nice idea. And results as I would had expected based on my previous experiences.

The results you achieved surprise me.  I would expect larger variances.

Great to have real world examples vs what the sales person says. Thanks!

The results you achieved surprise me.  I would expect larger variances.

If the sky is good at both base and rover the results will approach that of the total station.  

I assume the units are feet (not meters) since you mention 300’ intervals?

4 constellations?

@jon-payne That's some pretty fancy dirt if it needs to be better than a tenth. Bluetops are one thing but if it is just site dirt, I'd be curious what the circumstances were that a tenth mattered.

I assume the units are feet (not meters)

Yes. International Feet, not meters. GPS, GLONASS, & GALILEO. 

@lurker I'm not sure what part it was, but there is a large sewage lagoon thing being built to serve the residences being developed.  Maybe that concerns the earthwork guy more because it is an unusual build in this area and he doesn't want any problems being pointed his way.

IMO, it really doesn't matter too much what it was being graded.  When the contractor is better at measuring than the surveyor being paid to lay it out, it looks bad.

IMO, it really doesn't matter too much what it was being graded.  When the contractor is better at measuring than the surveyor being paid to lay it out, it looks bad.

Eh, I wouldn't say that it looks bad, or that the contractor is better at measuring. The contractor decided to use a micrometer where an elementary school ruler would do, while the surveyor understood the application from the start.

We set rough grade stakes and mark cuts/fills to the nearest tenth, which is plenty good enough for the typical application. If a client needs bridge girder tolerances on rough grading staking, they better say so when they call to ask for a proposal.

Long, long ago (in a city not far away) when I was a budding IM we had a party chief who didn't trust the Lenker and would only use a Philly rod.  His other quirk was that he wanted you to set up so your height of instrument was xxx.99 or xxx.00 with reference to the bench mark.  It was often awkward when a half of a foot was more convenient and comfortable.

(I think this belongs in the next post)

Nice. I did something similar on a job but it was wide open. I used rtk base rover and total station and ran some levels. The rtk beat total station as the distance increased over 400 ft up to 800 ft it was steady more error in a robot face 1 and face 2.  Now that same site at a different time when i had no good sat above 65 degrees elevation and close to overhead the rtk was not as good. On a select set of points. If you understand how the equipment was designed to work not all brands use the same approach. But all just about can achieve the results you need if you understand that and pay attention.  Nice very nice.  

IMO, it really doesn't matter too much what it was being graded.  When the contractor is better at measuring than the surveyor being paid to lay it out, it looks bad.

Eh, I wouldn't say that it looks bad, or that the contractor is better at measuring. The contractor decided to use a micrometer where an elementary school ruler would do, while the surveyor understood the application from the start.

We set rough grade stakes and mark cuts/fills to the nearest tenth, which is plenty good enough for the typical application. If a client needs bridge girder tolerances on rough grading staking, they better say so when they call to ask for a proposal.

This is exactly right.  It is all about being efficient in the field.  Back when I first started surveying I worked with a party chief that would lay out rough grading for roads with 10'+/- fill and cut.  He would use hubs WITH tacks.  He would stake out every 25' centerline on straightaways and 12.5' in curves.  If he staked 10+00 and 10+00.30 was the VPC, they BOTH got staked.  He had color schemes for everything.  VPC Red and White, VPI Red and Blue, VPT Red and Yellow, PC Orange and White, PI Orange and Blue, PC Orange and Yellow.  I later found out that the contractors knocked out everything with red on it, because they thought red meant no good...  

I then worked with my second party chief, and the same roads in the same conditions got 100' on straightaways, 50' on curves and if the PC or PT were within 10' of an even station, the even station did not get set but the PC and PT did.  The cut was nearest foot marked from the top of the stake.  Same roads got staked in 1/4 of the time and never had a single complaint.

With previous employers, I have built ENTIRE subdivisions with RTK Base/Rover all but the gravity sewer.  Site drains perfectly well, the roads are fine, not a single problem.  

Now, am I going to go out and lay out a runway with RTK?  No.  Well, maybe the rough cut =)  Am I going to lay out sewer running at half a percent with RTK?  No.  Do I need to break out a level on rough grading for a site?  No.

If that is what is required by the contractor, then they sure as heck better say that upfront.

These views are my own and not that of any current of former employers.

Very similar to the checks I did a couple of weeks ago for two sites. My conditions were better than yours I'm assuming. One project has some timber, the other is wide open. My "bad" RTK check was 0.04', most of the rest were .02' or .03', with a few at .01' and less. These were only checks, basically I hold the rod over the control point and hit collect letting the unit collect the point within a few seconds; no bipod, no observed control point. Tight observations are reserved for important points such as property corners. 

While you and I may think 0.1' is good enough for gough grade on a site, it is the contractor's perspective.  I was speaking with a foreman one time who was building a site to spec, and was constantly flirting with the the lower end of the grades - necessitating more surveying checks and fees.  I asked why they didn't shoot for the middle of the spec, and he told me that over a large site, evey 1/4" is thousands of dollars in hauling and dirt grading fees.

While I think the middle of a spec is the place to be, the contractors have a different perspective.  We stake to the tolerance that they need.

@chris-bouffard You would see greater variances if using VRS corrections over internet.  On site base over radio will give tighter results due to the lower latency.

"....he told me that over a large site, every 1/4" is thousands of dollars in hauling and dirt grading fees...."

True enough. But they cannot grade to 1/4" tolerances in any case. 

@toivo1037 

Unless someone explicitly requests it in writing, I'm going to stake design grades, not perspectives.

If a rough design grade is 150.0', and the contractor wants us to stake something other than 150.0' at a tolerance of other than 0.1', that needs to be spelled out and approved well in advance. I'm not going to be left holding the bag if and when the final lift doesn't bring the finished grade up to within spec and our own as-builts show it to be out of tolerance.

And if they are needing additional survey work to try and cut it as close as possible to tolerance, those additional fees (for us they would be classified as "extras & restakes" and billed as T&M since we already fulfilled that part of the staking contract) might well approach the amount of money they are trying to save.

But they cannot grade to 1/4" tolerances in any case.

For sure. Likely why they were needing all the extra survey work. Chasing that 0.01' above bare minimum required grade, wasting money in order to try and save money.

. Chasing that 0.01' above bare minimum required grade, wasting money in order to try and save money.

Dewey, Cheatem, and Howe have to eat too and make the boat payment college fund etc....