Infiltration Testing Techniques 2

[bucknell06]

In my area, I have been seeing a lot of infiltration-based stormwater management systems being put on site plans lately. I am curious to hear thoughts on A) the role of geotechnical engineers in the design of infiltration-based systems and B) what do you think are acceptable methods of determing design infiltration rates of native soils?

I have been looking at various state stormwater regulations and am pretty amazed at the differences in requirements for infiltration testing. On one end of the spectrum, I have seen allowances for percolation tests, default values based on USDA classification, and determination based on gradation analysis. On the other end, I have seen requirements for double ring infiltrometer testing at the base of the systems. Anyone have any experience with these various testing devices/procedures: constant head permeameters (aka "Amoozemeter"); Guelph permeameters; or Turf-Tec's portable, small-diameter infiltrometer?

 

[bpattengale]

Most regulatory bodies i.e cities, watershed districts, exc. in my area accept infiltration rates based on the soil classification from the unified classification system. If requested the geotechnical engineer will give you an acceptable infiltration rate based on the geotechnical investigation, but they usually default to the unified soil types. Our state (Minnesota) stormwater manual has the recommend infiltration rates for the soil types and that is generally accepted practice.

The infiltration rates can be much higher than the recommend rates thought so it may pay to get a double ring done if it may reduce the size of your system significantly. Around here a double ring costs around $1500 per test and you can't do the test until excavation has been done to the final grade of the pond or infiltration structure. For those reasons I haven't see a lot of double ring tests done.

When I design infiltration practices, I generally like to get at least two borings in the infiltration area, usually the owner will only pay for one, but you take what you can get. Then based on the boring logs and soil type at the proposed elevation of the infiltration practice use the recommended infiltration rates. The recommended rates are generally pretty conservative and we've yet to have a system that doesn't draw down in the design period.

Its also good to have a geotech onsite during the installation of the system so they can make sure that what the borings say is there is actually there. If the soil type differ significantly overexcavation, gravel drains, or total redesign of the system might be needed to acount for the actual soil conditions.

 

[bucknell06]

Great information, thanks. The company I work for has the double ring infiltrometer testing equipment, but we too find that it can be cost prohibitive to use. You generally need an excavator, which will be sitting around the vast majority of the time; you typically need two people; and you need a water supply. The permeameters I mentioned in the original post appear to allow you to make a small hand augered hole and test from the surface down. They require a lot less water and can easily be used by one person. I wonder if these are a decent alternative to using the conservative design values. I would assume that the permeameters would be less accurate than the double ring and that more tests would be needed.

The states that we work in allow for default design values based on USDA classification and I'm pretty sure that is how most of the systems are designed.

 

[howardoark]

Is there any reason you wouldn't just do a percolation test - the type of test you're required to do for septic systems?

I understand that if there's a rock ledge three feet bgs you may not like the full-scale results. But, assuming you have some idea of what the subsurface looks like, wouldn't that be enough?

The only time I see SDRI's being used is for landfill liners where the client can afford them.

 

[BigH]

Why can't you put in a drill rig, case the hole to various depths and carry out falling head permeability testing? This is described in detail by Hvorslev and also given in Fang's Foundation Engineering Handbook (Fig 1.19) among many other references. This used to be standard borehole data collection when I was in the field.

 

[bpattengale]

howardoark doing a percolation test as I understand it would still have the same drawbacks as a double ring. You would still have to excavate down to the finished grade of the infiltration practice.

BigH that sounds like a good idea. However I've never heard of that being done before, but then again I'm pretty young. Most of the borings I get reports for are for 8-inch hollow stem augers, what kind of cost difference are we talking if you start casing the bore hole?

 

[muuddfun]

We usually percolation test them, throw a 3 inch perf pipe down your holow stem, then do a falling head test on it.

 

[bucknell06]

Thanks for the replies.

BigH, I don't think there is any reason why you couldn't perform falling head permeability testing during the initial site investigation to obtain a preliminary value. We do this frequently as part of the normal procedure for site investigations if the client knows the location of a retention pond, etc and sees the value in the small extra cost. However, if the testing is requested after the initial site investigation has been performed or the layout or grading change, the owner doesn't always want to pay to mobolize a drill rig again. Also, if a state allows a civil engineer to look at a boring log or test pit and use default design values, that seems to be the way that it is generally done.

Bpattengale, as far as cost diffrence goes between cased holes and augered holes, it all depends on soil type, depth of hole, and how your driller's rates work. In my area, augers and casing should both be part of the driller's standard equipment as either one may be needed on any given site. Generally speaking, if your driller is working on a day rate basis, casing a hole MAY not be any more expensive than using augers. In certain subsurface conditions, casing is the best, and sometimes only option. In my experience, if footage rates are used, casing will be more expensive than augers to drill the same footage. There are lots of other variables to consider. Falling head perms take time, so the drillers will have some down time.

My understanding of perc tests is that they have been proven to be suscptible to huge varitions even in very small changes in locations. My own experience with perc tests is that an engineer familiar with the local soil conditions can observe a test pit and estimate an infiltration rate more accurately then a perc test performed in the field. However, I know that different states have different perc test methods. The type of perc tests I am familiar with don't necessarily involve an excavator, but do require a post hole digger and a marked-up grade stake.

Maybe my situation is unique, but I have seen new proposed regulations pushing the field methods indicated in my orignial post. The double ring is particularly good for verification testing at the base of the system during construction. I imagine the redesign would be pretty tough to do at such a late stage, but I have seen the verification testing requirement written into some regulations.

 

[BigH]

I carried out many investigations in the 70s and 80s throughout Canada and, as the geotechnical consultant, if we knew there was a need for permeability determinations, then we did the permeability tests. As shown in Hvorslev, there are many different "conditions" that could be used. As for the client - most investigations are preliminary at the start and if there are needs for more detailed investigations they are usually done - there is an old adage - you pay for it now or pay for it later. If he needs the data, he needs the data. The "classification" systems, in my view, are presumptive. If necessary go to the informaiton given in Terzaghi Peck Mesri where they use Kenny's work in deriving coefficient of permeability from grain size for a wide range of materials - not just medium single size sand (as per Hazen's original assumptions which most don't realize).

 

[fattdad]

Here's the field program that's being used around here (Central Virgina and to the north):

Drill a SPT boring to a depth that extends at least 4 ft below the design base grade of the infiltration basin. Do a hydrometer test on one of the more restrictive samples and correlate to USDA texture classes (i.e., Texture Class I, II, III, or IV) with the appropriate perc rate for each class). Advance a second boring just below the infiltraion base grade. Install a 4 or 5-in diameter PVC pipe to that grade. Install a column of water and allow a 24 hour soak. After the 24 hour soak, provide a 2-ft high water column atop the bottom of the pipe and measure the infiltration rate for four consecutive hourly readings. Report the individual values and the average along with the soil boring logs and hydrometer/texture class data.

Here's the stupidity of this entire process. The design is based on the water column being infiltrated in a 48-hour period. The popular thought (not mine) is that if you have an infiltration rate of 60 minutes per inch a 48-in water column will infiltrate fully in 2 days. That's why in many areas you have to document that the water table is more than 4 ft below the base grade of the infiltraiton bottom. Well if you are just infiltrating water from a 4-in diameter pipe that may just be the case. It just won't work like that if you have a 20 ft by 40 ft "cell" that has a 4 ft separation from the water table. What matters most is the horizontal permeability. At some point you just can't just keep pushing water down and it has to flow away. This may be hard to explain in typing, but this is the big problem in my mind.

Hope this helps.

f-d

¡papá gordo ain’t no madre flaca!

 

[bucknell06]

Thanks Fattdad, very informative. I have seen diagrams of that testing procedure (Fairfax County website, I believe?) and was hoping someone would have experience with it. I imagine that 24 hour soak and 4 hourly readings could add some significant extra time and cost to the project, especially if exploration would otherwise only require one day. I guess if it is a requirement then everyone would include it in their proposal.

Like you, I would tend to think that there are going to be other factors at play in a full scale system that are not replicated by a 4" diamter pipe. So my question to you is this: in a perfect world with no cost limitations, what testing do you think would be most appropriate?

 

[fattdad]

It's not that I feel the "Fairfax" method is inappropriate. I rather like it. That said, it's likely not enough. Figure that in many soils, the horizontal permeability is greater by a factor of 4 to 10. Figure also that the layer thickness of the most permeable "horizontal" layer may be just a few feet. So if you have a 30 ft by 30 ft infiltration basin all the water has to soak into this 900 sf area, but flow horizontally out of a 720 sf "layer surface". That's not too far off, but the horizontal flow gradient will not be as great.

Infiltration tests are primarily governed by a vertical flow gradient of unity. Irrespective of the "soaking" period, the Fairfax method is really showing infiltration witn a vertical flow gradient of 1.0. There is no way that the horizontal flow will be governed by a gradient of 1.0! It'll be much less.

So, to make a long story short(er), I'd try to get some fix on horizontal permeability, some fix on the thickness of the most permeable horizontal layer and then consider whether you can move all the water that you collect horizontally away from the infiltration "pond".

There are mounding equations, well equations, etc. to help with this.

f-d

¡papá gordo ain’t no madre flaca!