Strange Piezometer Behaviour

[starnosemole]

Hi all,

We've been working on a bridge construction project since 2004, in which a road bridge over a main railway line and a secondary spur line required some very in-depth analysis when it was discovered that a 100+ year-old sewer (combination of soil supported brick arch and pile supported cast iron pipe) existed in conjunction with an infilled inlet - the bottom of which (approximately 7 metres below ground surface) was a deposit of Peat up to 6 metres thick. In order to construct the large abutment fills, the sewer was re-routed, as well as a number of other utilities (including a 500mm diameter HP natural gas line). Wick drains were installed through surficial fill, Peat deposits (south of tracks), Clay/Silt deposits to underlying Silty Clay Till on approximately 10-foot centres and filling began.

A string of VW piezometers and two VW settlement gauges were installed below each abutment fill. The following was noted abut the piezometer installations:
- piezometers were installed on 2 metre centres with the shallowest being approximately 10 metres below original ground surface (ie - 10, 12, 14m sensor depths)
- piezometers were installed in sand lenses, separated by deposits of bentonite
- South abutment fill piezometers were installed into a large volume of water in the hollow-stem augers, whereas the North abutment fill piezometers were installed into a limited volume of water in the hollow-stem augers
- the augers on the North abutment fill removed a 3 metre long section of wick drain (wrapped around the augers)

Following the completion of abutment filling and surcharge operations, the piezometer readings peaked and began to fall similarly on both sides. The site was closed to construction activities. In March, approximately 3 months following the cessation, the upper piezometer on the north side began to rise. Around May, the middle piezometer on the north side began to behave similarly. By June, the upper piezometer and lowest piezometer were reporting similar pressure readings, and by September the three piezometers on the north side were all reporting similar pressure readings (within 2 per cent).

The piezometers on the south side have demonstrated a return to near pre-construction hydrostatic conditions, however, the north side piezometers now show divergent hydrostatic conditions, the upper piezometer reading a hydrostatic head of almost 2 metres into the abutment fill (around 2.5 times the original hydrostatic head).

The settlement gauges on both abutment approaches have behaved similarly.

In researching papers on piezometers, I have read that piezometers can demonstrate a pressure increase when they fail (giving readings "off the scale" though) and gas pressure trapped (ie: Methane) in the soil can produce erroneously high readings.

Has anybody seen similar piezometer(/piezometric?)behaviour on other projects?

Many thanks!
Steve

 

[dgillette]

Yes, I have. It is not real unusual for VW piezometers to go bad, and when they do, they usually show increases in u, until somebody decides the results are physically impossible, or at least implausible, and the VWs get deleted from the reading schedule. The instruments that I've seen were made such that higher pressure reduces the tension in the wire, so if the attachment of the wire slips or if something creeps, the reduced tension in the wire appears to be an increase in pressure. Another thing that can happen is that they aren't completely saturated when they are installed, and the zero point gets (messed) up. I would certainly be suspicious of a piezometer that reads much higher than one that's only 2 m away from it, absent a large supply of water. I wasn't sure from reading your message how many meters of head differential there are between the two piezometers, but it sounds like it would be more than 2 m, indicating a gradient greater than 1. Seems pretty high unless there is some unusual groundwater condition, or consolidation hasn't gotten very far yet. Check whether the settlement gauges indicate consolidation is nearly finished. If so, that is evidence that the VW is bad.

I haven't run into the methane issue before.

Interesting choice of handle, starnosemole, and considerably more creative than some others around here, e.g., dgillette. (Starnose mole is an endangered species as I recall - correct?) Is there a good story behind the handle?

 

[starnosemole]

Thanks dgillette,
It's possible that they weren't saturated on the north side, but what I find strange is that the upper piezos went in sequence to capture the lower piezos reading. Judging from the behaviour of the settlement gauges, and the instrumentation on the opposite side of the railway, I'd say something is effecting the upper gauges. As of the latest readings:

Piezometer Installation Elevation Piezometric Elevation
Upper 71.7m 86.3m
Middle 68.7m 83.1m
Lower 65.8m 80.3m

We expect that the original groundwater level would be around 77.2m. Original ground surface elevation is approximately 83.0m.

As for the Methane, I suspect that the decomposition of the peat may produce some gas, but not in great volumes. What I'm concerned about is the possibility that there may be gas pressure being applied by a different source.....say, a leak from the relocation of the HP gas pipeline? "Crazy talk" maybe....but maybe not....

..on the "handle:

Marvin The Starnose Mole Buys a Car sketch done by The Frantics was my favourite of theirs - they were on CBC radio and then on TV for 13 episodes here in Canada. (there's some of their stuff on You Tube like their most popular "Boot to the head" sketch:

http://www.youtube.com/watch?v=J5kGUW6M7W0

and their web site is at

http://www.thefrantics.com)

It cracks me up every time I watch the old VCR tape, when the salesman asks, "This dirt thing is pretty big for you, eh Marvin?" to which Marvin replied, "Dirt is my life". Seems very appropriate for a Consulting Geotechnical Engineer.

 

[dgillette]

Interesting. Each of the piezos is reading 14.5 +/- 0.1 m of PRESSURE head, but the TOTAL head gradient is about 1.0 between each pair. I don't have an easy explanation unless there is a source of perched water at/near the surface and it's seeping straight downward toward a water table.