Rockfill Dam FIlter Criteria

[moe333]

I am using filter criteria by Sherard and Dunnigan to design a filter for a clayey silt (52% fines) core in a Rockfill dam. The filter according to this criteria would have a D15 less than 0.7 mm, and less than 5% passing the #200 sieve. A typical concrete sand would fit this criteria. I assume the less than 5% passing the #200 sieve criteria is to reduce head within the core.

I would be concerned about water pooling in the filter since the zone downstream of the filter will be a silty sand and gravel (decomposed granite) which may have about 20% fines, and will be significantly less permeable than the concrete sand. I could include a blanket drain to tie the filter into the outer rock fill shell, but I wanted to avoid the use of a blanket drain if possible.

There will be several material zones downstream of the core filter. Each zone will be progressively more permeable, and coarse, until they reach the downstream rockfill shell (2 foot minus material). The way I interpret the Sherard and Dunnigan criteria, each of the downstream zone materials would need to have less than 5% passing the #200 sieve. This doesn't really make sense to me.

Any comments or suggestions regarding the 5% passing the #200 sieve criteria, or transition/filter criteria for the progressively more permeable downstream material zones.

Thanks

 

[dgillette]

You are correct. The 5% criterion is for permeability.

Could you put a chimney drain downstream of the filter? That way, the permeability of materials farther downstream is of little consequence, since the chimney drain would intercept all the seepage. If the core is built as designed from the presumably plastic material you have indicated, the amount of seepage through the core should be VERY small, and the filter would have to conduct almost no seepage at all except where there is a flaw of some sort (out-of-spec lifts, cracks at foundation irregularities, etc.). The chimney drain needs to be filter-compatible with the filter, but if you put it in, the gradations of the materials downstream are of little consequence because the water will be intercepted and drained off. Putting in the chimney means that you are not depending on the vertical permeability of the filter, which can be << the horiz perm because of the layering created by placement and compaction.

DRG

BTW - don't let them over-compact the filter. 80% RD is not better than 70% RD! Too much compaction creates a low-perm zone at the top of each lift, drastically reducing the vertical permeability

 

[moe333]

DRG,

So the chimney drain would outlet to the rockfill shell via a horizontal blanket drain at the base of the dam?

Do you think the DG downstream of the chimney drain need to be filter compatible (>5% fines)to the chimney drain?

Thanks for your comments

 

[dgillette]

The chimney should go somewhere, via either a blanket or collector pipes. The blanket is preferable because it won't crush, corrode, need repair, or be a home for muskrats. More importantly, the blanket drain can also pick up seepage coming up from below, out of the foundation, which is usually greater than seepage through the core, and filter it.

No, I don't think the DG really needs to be filter compatible with the drain, unless you think there will be significant water flowing in that direction (and the 5% criterion is just to minimize head loss - no significant flow -> no significant head loss). The chimney is intended to intercept the seepage and prevent it from going any farther (taking it down into the collector or blanket).

If DG is to be placed over a blanket, consider a transition zone (perhaps select DG) to keep infiltration and gravity from moving fines into the blanket.

Regards,
DRG

 

[BigH]

I acquiesce to David's expertise in dams, but might I also suggest that you have a look at Cam Kenney's work on filter criteria (referenced in Terzaghi, Peck and Mesri). Very good. Charles Ripley (Ripley Klohn Leonoff) always stated that good concrete sand would be excellent filter for any fine grained soil.

 

[dgillette]

We often try to make our filter specs fit C33 concrete sand, since that will filter almost any soil, and it's easier to find than custom blends. (I've only seen one exception where C33 did not meet Sherard-type filter criteria according to the numbers. We didn't actually run tests on it, so I don't know for sure that it wouldn't actually work.) Mr. Ripley was right on with that. One potential drawback is the limited permeability of sand that goes down to #100 sieve.

The one thing you have to watch out for with other criteria is gap-graded or "flat-graded" soils that aren't internally stable. That's why Sherard has you do the analysis on the minus-#4 fraction of the base soil.