In a message dated 17/05/2010, icarus@......... writes:

I called Dr. Haydar Al-Shukri today, and he told me the story of water in  
the vault.
_http://www.eas.slu.edu/Earthquake_Center/NM/UALR/ualr.pict.html_ 
(http://www.eas.slu.edu/Earthquake_Center/NM/UALR/ualr.pict.html)  

The vault was installed while he was working at St. Louis University, and  
he is currently at the University of Arkansas at Little Rock. The wires from 
 the vault to the building were encased in PVC pipes.  The seismometer was  
manufactured by Guralp. One day the instrument stopped working and they  
investigated as to why. After checking all the instrumentation in the  
building, they opened up the vault and it was filled with water. It  turned out 
that a tree had fallen down and they brought in heavy equipment to  remove the 
tree. Driving over the buried PVC pipe, the heavy vehicle had  fractured the 
PVC pipe........then, there was a heavy rain, and water leaked  into the 
broken PVC pipe which filled the vault with water. The vault was  emptied and 
the instrument forward to Guralp for  repair.



Hi Bob,
 
    One of my previous comments was that the cable pipe  seemed to be 
buried only about 6" deep. That would agree with the crush  damage. Drains and 
water pipes are normally ~18" below ground level for several  very good 
reasons.
    This suggests cold weather. PVC pipe normally  deforms and then springs 
back.
 
    This is still not very complementary about the  performance of the 
seismometer. It should be waterproof and floating! Guralp did  have some leakage 
problems with early seismometers.
    Quote: The CMG-40T is ''Self-contained in a  waterproof stainless steel 
case'' Umm?
 

If one were to build a similar setup as the U of  Arkansas did, what be the 
best installation of the connecting pipe  containing cables?  Perhaps 
V-shape or dip with the pipe,  allowing for drainage? Would some sort of a P-trap 
be applicable?  How  could one control or prevent condensation etc?

    On the site photo it looks as if the seismometer  pit is downslope from 
the buildings. The ground is shale, which is fairly water  resistant. The 
pit should have been provided with drainage and the cable  extended beyond 
the seismometer and looped back. 
    Threading cable through 18ft lengths of tube can be  quite difficult. I 
hung the tube down a well and fed the cable in from the top.  You can get 
reels of strip steel for threading electrical conduit. 
    The concrete being poured looks as if it has stones  in it. It should 
be cement and sand in equal proportions.
    You can't seal the vault due to air pressure  changes. 

I was a little concerned with their installation among  all those trees. 
Having lived in Arkansas in 2008, I had severe noise  level from the tree 
motions on windy days. (But, I suppose you have to go with  what you got! Not 
all installations are ideal.)

    I would expect the location to be severely effected  by wind noise from 
the large tall trees. A seismometer needs to be well clear of  any tree 
roots. This was supposed to be a professional quality installation, but  it 
seems to be 'a bit short' on both planning and construction. 
    I note that the seismometer site was moved about  290 x 180 ft back in 
2007. 
 
    Regards,
 
    Chris Chapman






In a message dated 17/05/2010, icarus@......... writes:

  
I called Dr. Haydar Al-Shukri today, and he told me the story of wa=
ter in=20
  the vault.
  
http://www.eas.slu.edu/Earthquake_Center/NM/UALR/ualr.pict.html<=
/A> 

  
The vault was installed while he was working at St. Louis Universit=
y, and=20
  he is currently at the University of Arkansas at Little Rock. The wires=
 from=20
  the vault to the building were encased in PVC pipes.  The seismomet=
er was=20
  manufactured by Guralp. One day the instrument stopped working and they=
=20
  investigated as to why. After checking all the instrumentation in the=20
  building, they opened up the vault and it was filled with water. It=
=20
  turned out that a tree had fallen down and they brought in heavy equipme=
nt to=20
  remove the tree. Driving over the buried PVC pipe, the heavy vehicl=
e had=20
  fractured the PVC pipe........then, there was a heavy rain, and water le=
aked=20
  into the broken PVC pipe which filled the vault with water. The vault wa=
s=20
  emptied and the instrument forward to Guralp for=20
repair.


Hi Bob,
 
    One of my previous comments was that the cabl=
e pipe=20
seemed to be buried only about 6" deep. That would agree with the cru=
sh=20
damage. Drains and water pipes are normally ~18" below ground level for se=
veral=20
very good reasons.
    This suggests cold weather. PVC pipe normally=
=20
deforms and then springs back.
 
    This is still not very complementary about th=
e=20
performance of the seismometer. It should be waterproof and floating! Gura=
lp did=20
have some leakage problems with early seismometers.
    Quote: The CMG-40T is ''Self-contained in a=
=20
waterproof stainless steel case'' Umm?
 

  
If one were to build a similar setup as the U of=
=20
  Arkansas did, what be the best installation of the connecting pipe=
=20
  containing cables?  Perhaps V-shape or dip with the=
 pipe,=20
  allowing for drainage? Would some sort of a P-trap be applicable? =
 How=20
  could one control or prevent condensation etc?
    On the site photo it looks as if the seismome=
ter=20
pit is downslope from the buildings. The ground is shale, which is fairly=
 water=20
resistant. The pit should have been provided with drainage and the cable=
=20
extended beyond the seismometer and looped back. 
    Threading cable through 18ft lengths of tube=
 can be=20
quite difficult. I hung the tube down a well and fed the cable in from the=
 top.=20
You can get reels of strip steel for threading electrical conduit. 
    The concrete being poured looks as if it has=
 stones=20
in it. It should be cement and sand in equal proportions.
    You can't seal the vault due to air pressure=
=20
changes. 

  
I was a little concerned with their installation=
 among=20
  all those trees. Having lived in Arkansas in 2008, I had severe noi=
se=20
  level from the tree motions on windy days. (But, I suppose you have to=
 go with=20
  what you got! Not all installations are ideal.)
    I would expect the location to be severely ef=
fected=20
by wind noise from the large tall trees. A seismometer needs to be well cl=
ear of=20
any tree roots. This was supposed to be a professional quality installatio=
n, but=20
it seems to be 'a bit short' on both planning and construction. 
    I note that the seismometer site was moved ab=
out=20
290 x 180 ft back in 2007. 
 
    Regards,
 
    Chris Chapman