ADDENDUM NO. 1 –
TIPS FOR DESIGNERS AND MSHA REVIEWERS
Regarding
Review of
Chapter 7, Seismic Design Stability and Deformation Analysis
Mine Safety and Health Administration (MSHA)
Engineering and Design Manual for
Coal Refuse Disposal Facilities
Advance Draft provided for Industry Review and Comment, December 2007
Prior to the days of fine coal cleaning circuits, many slurry impoundments contained deposits of non-plastic fine refuse composed largely of sand-size coal particles. At modern preparation plants, most of these coal particles are removed. The resulting fine refuse is typically a mixture of sand, silt, and clay that is cohesive. You may get some non-plastic fine refuse in the delta, but it often mixes with coarse refuse during pushout construction such that the underlying fine refuse is plastic. As construction progresses upstream, your dam can be founded on fine refuse that was deposited at the rear of the pond during the early stages of construction and often contains appreciable colloidal, clay-size particles.
Excess pore pressures can be significant during pushout construction over fine refuse deposited from modern coal preparation plants with fine coal cleaning circuits. That’s why the modified upstream construction (MUSC) method was developed so upstream construction can start early in a project and provide more time for pore pressures to dissipate. It's also why most combined refuse handling systems were abandoned in the 1980s because the development of pore pressures inhibit a machine from squeezing or sucking water out of clay (fine refuse filter cake) in a timely manner.
The presence of excess pore pressures in fine refuse beneath pushouts can trick testing methods like SPT and CPT/CPTu that were developed to test normally-consolidated deposits. Furthermore, laboratory testing methods have trouble modeling field conditions beneath pushouts. That’s why the conservative, simplified method of seismic stability assessment was developed for designing slurry impoundments with a residual phi = 4 degrees and a computed factor of safety in excess of 1.0. Strength parameters used in design can then be confirmed by in-situ residual vane shear testing performed during pushout construction.
As
described in the
previous recommendations,
continue to use the simplified method for seismic design and confirm your
strength parameters using in-situ vane shear testing performed beneath
pushout construction. Do not use the method proposed by MSHA for design
purposes, because it might create a potentially undesirable condition during
later stages.
CLICK ON IMAGE TO ENLARGE:
For
example, the draft MSHA manual presents a cross-section of a dam built by the upstream method following initial downstream development
as shown in Figure 16. If you drill from your initial pushout and obtain a
fine refuse sample that contains 70% sand-size particles and 30% silt and clay-size
particles with a PI of 5 (i.e. SC-SM type material), the draft design manual
recommends that you consider it to be sand and test it in the laboratory to
determine its undrained steady-state residual strength.
Figure 16
Don’t be
surprised if you perform a simple unconfined compression test of the material and
find that it has an unconfined compressive strength of 2500 psf as shown in
Figure 17, which means it is a plastic material with the consistency of stiff
clay. If you rely only on sus as determined in the laboratory,
you may be designing using the peak undrained strength and not the undrained
steady state residual strength, due to the limitations of laboratory triaxial
testing equipment.
Figure 17
If you had
designed the facility in Figure 16 using the simplified seismic stability
method with a residual phi angle of 4 degrees, results of your stability
analysis might have detected a potential problem as shown by the results in
Figure 18. For this example, you should consider starting your upstream
construction after Stage II, rather than waiting until after Stage III.
Lowering the coarse refuse/fine refuse contact elevation in the first pushout makes a significant difference in the factor of safety as shown by
the results in Figure 19. You can then verify the residual strength used in
design by performing in-situ vane shear testing during pushout construction
as shown by the results in
Figure 7.
Figure 18
Figure 19
-
When you perform SPT
testing from a pushout and encounter fine refuse with excess pore
pressures, you will measure low blow counts (N values). Don’t panic.
Bring the split spoon samples to the surface so excess pore pressures can
dissipate and perform pocket penetrometer testing. Obtain a few piston,
Shelby tube samples and perform unconfined compression testing in a lab as
shown by the results in Figure 17. Use the correlation shown in Figure 20
to determine the equivalent N values for the fine refuse deposit as shown
by the example in Figure 21.
-
CPT and CPTu probes will also be tricked
by fine refuse with excess pore pressure beneath pushouts. Unfortunately,
you don’t get a sample like in SPT testing, so you are on your own as to
what the results mean.
-
You may not be able to push the CPT and
CPTu probes through coal refuse after dissipation of excess pore
pressures. This equipment is expensive and there will be hell to pay if
the probe snaps off in the hole. Again, you are on your own here.
Figure 20
Figure 21
- Figure 22 shows some results of cyclic triaxial testing performed on fine refuse samples from the 1980s. Please excuse the metric units, but the conversion is 1 pascal equals 0.0209 psf. In these cases, a minimum of 100 feet of coarse refuse cover over the fine refuse for plants with fine coal cleaning circuits (i.e. cohesive fine refuse) and 150 feet of coarse refuse cover for plants without fine coal cleaning circuits (i.e. non-plastic fine refuse) provided conditions where pore pressures during an earthquake will be low, when construction pore pressures have dissipated and the material is normally consolidated.
Figure 22
-
Don’t forget that if you get sus
results from laboratory testing that are considerably higher than
your results from residual vane shear testing, please refer to
Figure 13. For cohesive
materials, laboratory testing procedures are strain-limited and may be
measuring the peak undrained shear strength and not the undrained steady
state residual strength. In such cases, ignore the laboratory results and
use the results of your in-situ residual vane shear testing.
-
If you must perform
laboratory testing of cohesive fine refuse samples to meet the
requirements of the draft MSHA design manual, then I suggest you consider
performing unconfined compression (qu) testing of piston,
Shelby tube samples. In a worst-case analysis, fine refuse will liquefy
during the BFE, resulting in an effective stress of zero. Although overly
conservative, the unconfined compression test is a simple laboratory
method to simulate such a condition. Figure 23 shows vertical effective
stress vs. shear strength (i.e. qu/2) data for undisturbed fine
refuse samples recovered from sixteen (16) sites compared to the design strength used
in the simplified seismic stability method. Even this overly
conservative method justifies using a residual phi of 4 degrees for design
purposes.
Figure 23
-
In going through this process, remember
what an MSHA reviewer said in the 1980s, “It’s not the coarse refuse or
the fine refuse that’s the problem, it’s the water”. You can deal with
the water in coarse refuse and non-plastic fines by providing drains and
adequate cover for rapid consolidation and confinement to avoid hinges,
but the water in cohesive, low permeability fine refuse is a different
animal. Load your fine refuse deposit with an initial pushout as early as
possible to allow more time for consolidation and pore pressure
dissipation. The installation of wick drains can help, but time is the
key. If you waited too long to start upstream construction, you’re on
your own.