Expert: Robert Cummings, P.E. - 3/4/2008

Question
QUESTION: Hi Robert,
Behind my house I have a 1:1 slope which since the loss of three large oak trees from the top of the slope I have had three superficial mudslides, the last two occurring with an erosion control blanket in place. Two local soils engineers have suggested That I remediate the problem by putting an interceptor drain system in front of my house and drain it to a seasonal creek at the slope toe. Access to the slope is not possible for a vehicle so a retention system on the slope would have to be done manually. My question is whether there is any way of predicting that dewatering the front of my house will protect the slope since I know that the crawl space below my house is perectly dry?

ANSWER: Because your crawl space is dry, this sounds more like an erosion problem than a geotechnical slope stability problem stemming from a high ground water table, but it doesn't have to be so.  Therefore, please clarify:

Your house is at the top of the slope, right?  How far from the crest?
How tall is the slope?
What part of the country is this?  About how much annual rainfall do you get and how is it distributed through the year?
1:1 seems very steep for a natural slope in soil.  Is the substrate bedrock with only a superficial soil mantle?
You seem to feel that the oak trees stabilized and partially dewatered the slope (which they may very well have).  What, in your opinion, killed the oak trees, and is there significant vegetation remaining on the slope?
By an "interceptor drain system" do you mean swales or lined ditches to intercept surface runoff, or do you mean french drains designed to intercept subsurface water?
These "superficial" mudslides -- what thickness and width of material was involved?

Without knowing answers to these questions, I can offer that if we are talking about surface erosion percolating into loose soils and causing mudslides, a properly designed and lined ditch system will probably work.  (I wouldn't call it "dewatering", though...)  You will probably still need slope protection.  There are all kinds of erosion control blankets out there and some of them are not very robust.  You may need to upgrade.

---------- FOLLOW-UP ----------

QUESTION: 1. The house is at the top of the slope.
2. The aorigalgrade beam and pier foundation is set back about 10 feet. A 15 yr old deck addition also with a grade beam pier foundationcomes to within a few feet of a 5 foot scarp at the top of the slope. The piers range in depth from about 18-22ft under the addition to 28ft under the house.
4. I live in the SF Bay Area and we get 20-30in of rain in the winter.

5. There has not been prior erosion/ mudslide activity in the 35 years I have lived in the house however we are at the foot of a valley and water problems are widespread including both neighbors.

3. It is about 28 fet down to a seasonal creek/bedrock.

4. When the piers for the addition were dug the geotechnical engineer approved it accting soils condition as satisfactory. Considerable ground water was encountered. One bore had to be abandoned and moved because of excessive ground water.

A very experienced soils engineer teo years ago after the first mudslide suggested a trench drain system in front of the house with drain connections  extending under the to the slope toe. In preparation he had a drilling company here planning to drill 2 bore holes on each side of the front of the house. The first borehole reached bedrock at 40 ft!! much lower that the creek bed behind the house. He  decided that a second bore hole didn't need to be drilled. He said he could see water on the bedrock flowing toward the creek behind my house but my vision couldn't confirm this He observed the bore hole for a few hours and said the ground water table was about 5 feet and his conclusion was that we should lower the ground water table. I don't know if this was an accurate way of measuring the water table
. I thought this could just be water slowly coming up the hole from ah aquifer.

5. We never had problems with the slope until we had to remove 4 oak trees from the crest which were afflicted with sudden oak death. he mud slide area was toatlly denuded. I would say the area is abou 600 sq. ft.

I mean a deep, 6ft or more trenched drain placed across the front of my house which can be connected to two drains guided by "directional drilling" to the toe of the slope.

5. ! would say the mudlides have each been afoot or so in tickness as even after 3 slides  there is hardly an noticeable difference in ground level between  involved and uninvolved slope.

5. You are not alone in trying to implicate surface wateras the causative agent. I don't however know how the rainwater is getting under the bblankets which were of good quality (North American Green). In fact I have just emailed and recieved a reply abot this from their local representive who sent it to his technical reprentative. The erosion control specialist who installed the second coir blanket was here after the recurrent slide. He thought there was a little recession aalong the grade beams realed to ground water slightly lowering the crest.

4. As to whether this is classifiable as surface erosion or a mudslide I would strongly favor the latter as it occurs suddenly during a rain and the mud ends up as bullocks or berms in midslope and at the toe.

5. If the drain system would work I of course would put a new blanket on the slope. The drain and mattress solution if viable would be a much more economical repair than soil nails and shotcrete or blanket which would be the alternative.

ANSWER: I have been considering this, especially the part about the shallow (5 ft) ground water table.  The geotech engineer said you should lower the ground water table (by installing horizontal drains, probably) based on his observation that a water rise in his borehole came within 5 ft of the surface.  (This by the way is an important point.  If water does, as you put it, "come slowly up from an aquifer", that means the aquifer pressure can drive slope instability and is therefore of concern.  It is water pressure, not the mere presence of water, that causes geotechnical instability.)  What is puzzling is that if a ground water table at 5 ft is responsible for slope instability in a 28 ft high slope, the instability would most likely be expressed as a deeper seated failure rather than a superficial series of 1 ft slides.  There is the less likely possibility that there is no aquifer system on your slope, in which case the "water table" was nothing more than the height to which seepage from a zone about 5 ft deep filled the boring while the geotechnical engineer waited.  If that occurred then he was mistaken in his interpretation.

What may be happening is that the soils are not so unstable that the deeper failure occurs, but the incidences of rainfall cause localized rises in ground water that issue from the slope face and cause the superficial mudslides.  There may be some liquefaction effects from small earth tremors, considering your area.  Formerly, the oak trees absorbed the water and prevented the buildup of excess soil pore pressures. If you are not seeing surface erosion than that is probably what is happening.  The interceptor drain is a relatively cheap alternative in case infiltration on your property is what is causing these transient water table fluctuations.  If the water may be coming from deeper or offsite locations, the interceptor drains obviously will not work and you will be faced with something more robust on the slope itself.

I would recommend you have the slope re-examined by a geotechnical engineer to be sure you are not at risk of a more general slide.  I would also, if I were you, plant some fast-growing ground cover on the slope, and some trees.  

You can consider upgrading the erosion control fabric so that the slope will remain more stable while the vegetation takes hold.  However it's a challenging application for just erosion control fabric.  You should consider a heavily-reinforced excelsior material -- not coir which breaks down in a season or two -- and stake it at least 2 ft deep all over the slope.  

If a geotechnical analysis indicates that deeper seated failures are a possibility, you need slope reinforcement or horizontal drains.  If you don't you are taking a chance that the slope will fail and take your deck piers with it because I will bet they are not designed for the lateral forces stemming from rapid downslope movement.  I assume you do not have access to the creek bed for drilling equipment so all work would have to be done from the top.  There is a firm that does launched soil nails that seem like an appropriate alternative if you need more general stabilization.   Take a look at www.soilnaillauncher.com.  The slope can then be covered with TECCO mesh (www.geobrugg.com) bolted to the soil nails and you will never have a failure again.

---------- FOLLOW-UP ----------

QUESTION: There is one lingering question which remains after your excellent insightful answers. Just to be sure of the ground water level in front of the house should we drill another bore hole and measure the water table correctly? Absent this we could find ourseelf excavating to 6'  or so with no groundwater. Doubtful but possible.  I think further deeper digging would require shoring considerably adding to the cost and how deep would we go? Remember we don't know what the ground water level is on the slope. The bore hole is at least 50 feet from the top of the slope!Can you also determine thhe groundwater level on the disturbed slope?

Thanks,

Mel

Answer
The intent of the intercept would be to remove infiltration before it becomes ground water. You are right -- trying to intercept ground water in a trench 6 ft deep would be a total waste of time.  If the ground water table was originally at 5 ft theoretically resulting in slope instability, lowering it to 6 ft would do practically nothing to enhance stability.

The idea of drilling another hole is tempting, but complete it as a cheap-and-easy monitoring well as long as you are at it.  These holes (when in soil) are normally drilled with a 6-5/8 inch hollow stem auger or a 4 inch flight auger.  You could drill it as close as possible to the slope crest and complete it by inserting slotted 2 inch PVC casing with a plug in the end, then backfill the hole to within 4 ft or so of the surface with clean -1/2 inch gravel or pea gravel, and cap the hole with a slurry of sand mixed with a little Portland cement to keep surface water from invading.  Leave the casing sticking up a foot or so and cap it to keep debris from entering.  Wouldn't take much gravel.  Then you can measure the ground water whenever you want -- dry periods, during and after storms, etc.  An alternative to slotted PVC casing would be to use solid Schedule 40 PVC and drill a bunch of 1/4 inch holes in the bottom 10 ft, say 1 each 6 inches at all 4 quadrants.  Wrap the casing in burlap or fiberglass window screen so that water can enter but the holes and casing are protected from plugging up (they eventually will but this should last a while).  The water will accumulate in the gravel and rise and fall in the casing according to whatever its static top elevation is.

You can get a rough idea of the phreatic surface by assuming that the water table makes a gentle convex-upward arc from the level in the monitor well to the level of the creek bed.  You would need a rough cross section of the slope surface from the creek to the well in order to visualize this.

If you don't see springs on the slope surface then the ground water table is not very close to it.  If water under any pressure at all can migrate to a free surface, it will.