Setting temporary points for settlement monitoring

This is always a strange one - the residents should be thankful that somebody is going to the trouble of checking their properties are safe, but they never are. Destroying the monitor marks doesn't make the properties any safer or stop the job; it just leaves the residents winding themselves up into a state of paranoia.
If you are looking to monitor settlement, rather than any horizontal movement then resecting the instrument position from 3 or 4 points in safe locations should be adequate.
For "marks" on the property it is better to use any identifiable "feature" and photograph it if possible so you can be sure it is always the same point you observe. If they have porches there will be points such as corners of door frame, joint in concrete, visible stone in concrete, nail in wood, etc. I'd be very surprisd if you couldn't find at least two identifiable points on each porch. Pencil marks can easily be erased, or worse - redrawn. If you can measure from nearby points, as you say, then you can probably use one or more of those as targets.
As you are looking for changes in heights, so long as the marks are more or less the same height as the instrument you could get away without an accurate distance - the change in VA (correct for instrument height) will identify a vertical movement.. If you can get all the points with a non-contact measurement then that is better.

WARNING - be careful to take a close look at the damp=proof course level. I've seen the main structure of a dwelling stay rigid whilst part of the foundations drop away beneath the damp proof course. If all your points are above that level then you wouldn't record any movement until BANG when the foundation gives up.

Retro reflective targets is how I would do it I would be very careful about puting a bench near the area to be monitored. Let's say you have a 3" gun you would need to keep the sight length around 300'. I would put a target 300' in the opposite direction of the area being monitored, preferably on a bridge abutment or something with a pile driven footing. The very least stable thing I would use is a light pole or power pole to mount a target. If none of those things existed I would go with a Feno rod.

0.005'? I think you need to use a digital level, run a closed loop from a stable point far enough away as to be unaffected by construction. Also, as stated by others above, I would not trust a porch to be as stable as the house, unless it has part of the house foundation underneath it. My porches, for example, are not connected physically to the house, they are simply concrete slab on a stone base.

We use a yellow paint stick when trying to be unobtrusive, and just place a dot (YPM=yellow paint mark)

Our goto was a yellow paint stick patch about the size of a gammon reel with a "+" marked on it with a #2 pencil.
When it was over 1,000 feet we would put up a square board that was painted red & white with lines from a black marker to set the cross hairs on and looked like the face of a gammon reel.
Our biggest board was 4ft x 4ft so we could see it from the other end of the main baseline.

It's an N3 job. Leave no trace? No way. A discreet permanent mark.

I ran a multi year job on 12 houses for the exact same reason. We used class-II procedures ever 30 days for 30 months. And with 40-50 marks, many didn't survive. Then after the project was over, we had to consider annual, or unscheduled future checks. 1" sq of stainless steel epoxied to structure, with a fine scribe line. Because if litigation returns, a second engineer/surveyor may come in. And you want agreement.

I agree that 0.005' is a job for a first order level, either digital or something like an N3.

On the house monitoring project a surveyor from the construction "couldn't detect any movement". So the city called us in.

Each point was measured better than 0.1mm. And over time a velocity was seen, i.e.; some pts were fast, slow, and stable. After several months, the movers showed predictable change. Even at a few 0.1mm/mo to several mm/mo. The line of stability v sink was found. And that's when repairs began. And repairs had to be shown as stable. It takes time and you cannot afford to lose pts along the way.

0.005 ft may be too noisy to see short term sink. A sink if 0.005' may takes weeks or longer to even show up. And if your not certain to better than min spec, you might mislead the engineers. And some points may accelerate or slow. 0.1mm standard err is the stuff of lawsuits. And if it's litigation, it's an expensive problem. And an expensive task.

And is your client the complainant or defendant? You might get called by the other side to give deposition.

Pencil marks? No.

The construction company may be on the hook for 10+ yr. People selling a couple of years down the road may have a problem if there's a cloud over the stability. So permanent and discreet.

Another comment...we do settlement observations on 39 dams. They points monitored are bolts grouted into drill holes in each monolith (mainly set in the early 70's). For horizontal there is a punch mark. For vertical, it would SEEM to be the top of the bolt, right? Not always. Sometimes these bolts are slightly recessed. So in reality you are monitoring the elevation of the concrete around the bolt in those cases. On one project this year for some unknown reason the crew (I was not out there) decided to use a spacer on most of the points, because they were almost all slightly recessed. This was on the newest project, built in the late 80's, so not the same people who set all of the rest. What is a spacer? It is a small precision dimensioned magnet that is 20 mm high, and about 16 mm in diameter. You place it on the bottom of the invar rod, and it is then placed on the mark. Of course, to properly use it you put it on the backsight of the last turn before the mark, then use it for both FS and BS on the mark, and then the FS on the next turn, so it all cancels out and it does not matter if it is exactly 20 mm or not (actually, I just measured 20.04 mm with a micrometer). The main use is on disks that are recessed below the concrete, because invar rods have a wide base that won't fit down on some of the disks if they are recessed.

Anyway, so they ran it and used the spacer on MOST of the pins, but not on the disks (4) as they were not recessed. But of course this did not match any of the previous surveys, which were actually on the concrete and not on the surface of the pin. I had to go back and rerun the entire settlement survey.

Larry Scott, post: 413280, member: 8766 wrote:
Each point was measured better than 0.1mm.

What are instrument are you using that provides and accuracy of 0.1mm? That is 0.00033' or 1/256" or the thickness of a piece of paper.

You can get that precision on a single point, using a high end digital level or an N3, etc with micrometer, but not that accuracy point to point to point.

We run a lot of settlement surveys, the best you are going to get over a series of points covering a decent size area is 1 to 2 mm, and that is with invar rods, balanced sights, etc. Maybe inside the SLAC tunnels, controlled environment, etc you can get better, but in the real world, outside, those are realistic results.

John Hamilton, post: 413291, member: 640 wrote: You can get that precision on a single point, using a high end digital level or an N3, etc with micrometer, but not that accuracy point to point to point.

We run a lot of settlement surveys, the best you are going to get over a series of points covering a decent size area is 1 to 2 mm, and that is with invar rods, balanced sights, etc. Maybe inside the SLAC tunnels, controlled environment, etc you can get better, but in the real world, outside, those are realistic results.

That was kind of my point. Accuracy, precision and the display results on the digital level are all very different beasts. 1.5mm accuracy is achievable with rigorous procedures; high quality, calibrated equipment; and optimal environmental conditions. I just doubt you could get 15 time better accuracy with a differential level.

John Putnam, post: 413305, member: 1188 wrote: That was kind of my point. Accuracy, precision and the display results on the digital level are all very different beasts. 1.5mm accuracy is achievable with rigorous procedures; high quality, calibrated equipment; and optimal environmental conditions. I just doubt you could get 15 time better accuracy with a differential level.

I think the only way you could get .1 mm is a single setup taking repeated shots on a point, without changing the setup. Or maybe transferring elevation from one point to another that is close by. Once you start adding in turns, and turning points, and different setups, the accuracy will begin to degrade.

John Hamilton, post: 413291, member: 640 wrote: You can get that precision on a single point, using a high end digital level or an N3, etc with micrometer, but not that accuracy point to point to point.

We run a lot of settlement surveys, the best you are going to get over a series of points covering a decent size area is 1 to 2 mm, and that is with invar rods, balanced sights, etc. Maybe inside the SLAC tunnels, controlled environment, etc you can get better, but in the real world, outside, those are realistic results.

With an N3, or Jena, micrometer, dual pattern Kern Invar rods, double simultaneous reciprocal, I've run loops at mm closure.

However ... A good Invar rod is dual pattern and 1/2 cm. And the micrometer reads to 0.1 of the pattern, and estimating to 0.01 of the 1/2 cm pattern. So there is averaging. (Which is only a reading.) However in structure monitoring you're typically dealing with very short sights and minimal number of turns. And very very important: balanced sight distances. BS to FS should be balanced to a meter. And that'll make a significant improvement on repeatability.

Sub mm isn't an overstatement.

As for the original question, micrometer reading N3 and permanent marks.

Just sayin'
Philly rod, pencil marks, isn't good enough. Procedure, marks, and the best rod instrument pair you find. Sight length short and balanced. Permanent turning points like a pk epoxied in concrete drill holes. And monitor the turn pts as well as the marks on structure, and one can easily do better than 0.005', and should.

John Hamilton, post: 413286, member: 640 wrote: What is a spacer? It is a small precision dimensioned magnet that is 20 mm high, and about 16 mm in diameter.

I used a spacer for the first time a couple of months ago. There was one mark in my network that was carelessly set such that the concrete protruded slightly above the disk, preventing the rod from setting on the disk. I used a cylindrical hardware store magnet, it was nicely cut, with smooth and parallel ends (as parallel as I was able to measure, anyway). I measured it with a vernier caliper, it came out to 0.042-foot after converting.

We only needed it on that one mark, so for procedure we used verbal "spacer on" and "spacer off" calls. For calculations I copied the appropriate L records in the Star*Net file, saved the originals as comment lines, and manually changed the height difference values on the two affected records to account for the spacer. I also added comments describing why those records were altered.

Since this was a 3D project I also had GPS and total station measurements on the mark, which provided a blunder check on the spacer process.

I don't think there is a significant difference in results attainable under normal conditions with a good digital level (we use a Dini12) versus an N3. The N3 is definitely superior where there are vibrations or high wind, because it does not have a compensator. But it takes some skill to properly use the micrometer, and there are not that many around. I saw one on Kara's web page for $7900.

Like I said, we probably run more precise levels than 99% of surveyors, and attention to detail is important. Typical closures are sub millimeter, but I am not deluding myself into thinking that all of the points are accurate to sub millimeter. 2 mm is our re-run criteria on dams, and the loops are usually about 1 km, although we do have two dams that are 2 km one way across (making a 4 km loop).

Here are some results for a project I just submitted to NGS. It is a second order class I leveling of an area that we relevel every year, and submit to NGS every 5 year. Part of the area is slowly subsiding due to a salt mine 2000' down, a few mm per year. All values are in meters, except distances are in km. Some sections are only run one way because they are outside of the subsidence area and they check against published, meaning they haven't moved. Allowable is second order class I allowable misclosure (0.006 m * sqrt(km))

Per the attached, Trimble specs their high end DiNi at 0.3mm / km of double run levels, and the least count on it is 0.01mm / 0.001" / 0.0001'. You can also set it to take multiple measurements with a maximum deviation of whatever value you're happy with. For sure, error will creep in as John said, and he has a lot more experience with that kind of work than I do. But 0.005' is about 1.5mm; that should be attainable with good procedures. Obviously, good turning points and balanced turns are both necessities.

Yes, it specs at 0.3 mm/km, but reality in the field is sometimes a bit different. We use 3 m rods with struts, turning pins or plates, balance our shots, do a Kukkimaki every day, etc. The project I mention above is on the shore of one of the Great Lakes, and the wind is often a major factor. The struts help keep the rod steady, but it definitely affects the level. We set the standard deviation limit to 0.1 mm. That is easy to get on a calm, cloudy day. Wind, sun, truck traffic, longer sights all mess with that.

John Hamilton, post: 413357, member: 640 wrote: I don't think there is a significant difference in results attainable under normal conditions with a good digital level (we use a Dini12) versus an N3. The N3 is definitely superior where there are vibrations or high wind, because it does not have a compensator. But it takes some skill to properly use the micrometer, and there are not that many around. I saw one on Kara's web page for $7900.

Like I said, we probably run more precise levels than 99% of surveyors, and attention to detail is important. Typical closures are sub millimeter, but I am not deluding myself into thinking that all of the points are accurate to sub millimeter. 2 mm is our re-run criteria on dams, and the loops are usually about 1 km, although we do have two dams that are 2 km one way across (making a 4 km loop).

Here are some results for a project I just submitted to NGS. It is a second order class I leveling of an area that we relevel every year, and submit to NGS every 5 year. Part of the area is slowly subsiding due to a salt mine 2000' down, a few mm per year. All values are in meters, except distances are in km. Some sections are only run one way because they are outside of the subsidence area and they check against published, meaning they haven't moved. Allowable is second order class I allowable misclosure (0.006 m * sqrt(km))

Given the frequency of your sub millimeter differences, F/B over short distances, and half of all runs have differences 0.0000-0.0006 mm. I think over short runs for structure monitoring, 1.5 mm (0.005 ft) is loose.

IMO, 0.005 ft in structure monitoring may not show what he needs, and much tighter is not over the top procedures.

So, 2 thenths mm, std err, isn't rare.