Would love to see that "bucking in" stuff input to least squares. They are monitoring "something" but it sounds like a bunch of one-third-giraffe scale monkey business.
?ÿ
I was getting reference factors of 0.1 to 0.2 when using standard errors that I know for a fact are correct for our instrumentation.
For comparison, a tight network run by a crack crew with recently-adjusted instrumentation typically runs in the 0.75-0.90 region. Usually right on the edge of failing the chi-squared test. Our less-precise work might be in the 1.2-1.5 range.
Neither would raise red flags, but 0.2? I don't buy it.
It just smells. And it's really hard to explain that to folks who have no idea what I am talking about.
The part about "staking to" the monitor points but not recording anything, combined with sliding the instrument around until it has zero residuals from the record control, all sound like a method to report minimal movement every time, rather than actually monitor anything.
?ÿ
The procedure described in the original post sounds like a lot of work that fails to achieve the desired goal.?ÿ I'd hate to have to defend it in court.
I have a small monitoring project that's going on 4 years now.?ÿ Nothing real critical -- it's a storm drain pump station -- so I'm happy with 2-sigma vertical errors less than 0.02' (they mostly run between 0.005' and 0.015').?ÿ Monitoring frequency varies; sometimes my client wants me out there every month, sometimes every 3 months.?ÿ
Everything is shot from one setup, but there was no convenient way to set a stable point there, so I treat it as a new point every monitoring event, checking to 3 offsite stable points at the start and end of each event.?ÿ Residuals to the stable points run between zero and 0.006'.?ÿ
I tape the HI of the setup each time just to get in the ballpark, but the coordinates for the setup point stay remarkably consistent despite its unreliability (it's a cup tack in a pavement headerboard) and the none-too-rigorous nature of the HI measurement.?ÿ The point coordinates have changed horizontally over the years by less than 0.02'; the vertical is sloppier as might be expected, showing as much as 0.03' difference.?ÿ It's hard to say how much of that is point stability, and how much is setup error, but it doesn't affect the monitoring either way.
I agree that the monitoring procedure is not ideal or what I'd do, but that's the way it was setup.?ÿ Changing the procedure midstream is not necessarily a better option.?ÿ
I would have done as asked and simply written in my journal that I did not agree with the process but felt that reinventing the procedure would cause greater harm.?ÿ If you're seriously concerned about professional liability, take a picture of your written objection next to a newspaper with a prominent date and there will be little chance of you getting popped in court for gross negligence.
?ÿ
I agree that the monitoring procedure is not ideal or what I'd do, but that's the way it was setup.?ÿ Changing the procedure midstream is not necessarily a better option.?ÿ
I would have done as asked and simply written in my journal that I did not agree with the process but felt that reinventing the procedure would cause greater harm.?ÿ If you're seriously concerned about professional liability, take a picture of your written objection next to a newspaper with a prominent date and there will be little chance of you getting popped in court for gross negligence.
?ÿ
A sad reflection on the current legal atmosphere--almost sounds like you are a hostage.
I agree that the monitoring procedure is not ideal or what I'd do, but that's the way it was setup.?ÿ Changing the procedure midstream is not necessarily a better option.
I should have made it clear in my earlier posts. This was brought to me literally the day after they started their "baseline" (using that term very loosely) "measurements" (also using that term loosely).
I would have done as asked and simply written in my journal that I did not agree with the process but felt that reinventing the procedure would cause greater harm.
I'm also not the crew chief on this - I'm a licensee who was asked to retroactively design processing and analysis procedures to fit their dubious measurement scheme, because the PM and associated leadership failed to do their jobs in the first place.
Reinventing the procedure could have been done. There was ample time to do it right, which is why I brought it up in the first place. And even if they were already well into the campaign, it would be one day's work to come up with a measurement scheme, properly observe the network and lock down both reference and monitoring points for future analysis.
Starting from scratch and doing it right would still be better from a public safety, liability, and best practices standpoint. I doubt that anything will go drastically wrong, but the small chance that it might is why we are getting paid good money to develop rock-solid procedures and take on a lot of liability.
@jaccen?ÿ
Well, I'm not likely to stick around after this incident, that's for sure. We blow a lot of money on lip service to our "health & safety culture" so we can put the sales pitch to our clients, but when it comes to actual public safety suddenly there's no time or money to do it right.
This was brought to me literally the day after they started their "baseline measurements" .
I worked for several years for one of the largest Portland area outfits, which had this management style:
1. Send crew to job with minimal instructions to set up program.?ÿ
2. Allow them to flounder around for several weeks until the budget is used up, or nearly so.?ÿ
3. Assign job to office LS to prepare map/report and sign off.?ÿ?ÿ
4. Criticize LS product for not following policy.
5. Invent policy du jour.
6. Bitch to LS about missed budget.
I've worked in several offices since then and find that this is fairly common. Even guys who railed against such management style and go into business for themselves end up doing it.
Thanks to everyone! Reasonably new to monitoring, and this discussion helped out a lot, even 3.5 years later.
My procedure would be to resect the instrument position each time from 3 or more points that are well offsite. If I was to set up?ÿ the instrument over a particular monument each time, and run this resection, I'd get a slightly different set of numbers for the instrument each time even though it was over the same point.?ÿ But it should be within the expected tolerance.?ÿ
This is the procedure I've been using at a drainage lift station structure that I've been monitoring for the last 8 years. It's a small site, about 140'x70'.
The image below is a closeup of the adjusted observation station positions, which is simply a cup tack in a headerboard. I set up over this point each time, measure the HI with a tape, then take measurements to the 3 stable control points, all of which are outside the settlement area. The spread of these positions is about 0.044'N by 0.008'E by 0.036'Z. The only reason I setup over the same instrument point each time is to ensure a clear line of sight to each of the monitoring points, as there's a lot of piping in the way, but it's been remarkably stable considering its nature.
P.S. The monitored structure hasn't settled a lot, less than 0.1' in the 8 years I've been involved. However, that's been enough to crack some pipe joints, and the contractor has had to pay for monitoring and repairs as a result.
Somehow I missed this thread when it was started, just saw it for the first time today.
As someone who has been doing monitoring surveys (over 600) for 45 years (since 1980), I consider the procedure described as LUDICROUS.
Here is the latest result for a point we monitor from that we know is unstable, but is the only place available. We setup on the mark every time. It is not part of the monitoring. It is a relatively narrow freestanding wall in the river, and has a clear seasonal (temperature) signal to it. It is not constrained in the adjustment. We do this survey monthly, and the next survey is tomorrow. The vertical axis is the offset of the point from a reference line which is the design alignment of the (failing) sheet pile wall being monitored.
My procedure would be to resect the instrument position each time from 3 or more points that are well offsite......
I should add that most of the monitoring I've done is on structures adjacent to construction sites being developed. In such a situation there is zero chance that I'd be able to establish a point to set up over that would be intervisible with the monitored points every visit. Too much stuff going on and being stockpiled.
I resect the instrument position for all types of work about 95% of the time. I set my control accordingly.
I've been monitoring dams and hydro facilities for a long time. I have three rules. First, keep it simple. Second, screw it up exactly the same way every time. Third, get enough data to test every piece of data three times. Hasn't failed me yet...