First, I take it that you have done the Lab Proctor to get the MDD values and OMC.  WRT to your second point, common sense doesn't tell me that you would expect only 40% relative compaction.  Look at the Lab curve and take the moisture content at compaction.  Then compare it to the ZAV line.  This will tell you what you likely compaction range is - I have presumed the pumping soil was, in fact, compacted in a proper fashion.  In fine grained soils (especially silts) they do start to pump under added loadings (passes of roller) if they become saturated after initial compaction.  I have seen road bases pump because they rolled to wet of optimum, the fines came up with the water during vibration of compaction, the water evaporated and the fines formed a thin film that wouldn't let more water escape - therefore pumping.
  WRT to your first point, again check out the Lab proctor, in situ moisture content and ZAV line. I would, though, wonder why you are trying to compact "organic" soil?  Are you actually using this for fill?  Or did you take out the NG to check on 'in situ' density?
  Also remember (with more granular soil) that you get something like 80 to 85% "compaction" just be dumping it out of the truck.  So your 40% expectations seem unreasonable to me.
  Let us know more after you have checked the points I raised.

Breezeear,
First, you didn't make it very clear if your dredgings are granular or more fine-grained, but I have the following comments;

I wonder if you might be confusing relative density and relative compaction with your "40%" estimate.  I've seen that confused by many.  As BigH said, you can expect to get 80 to 85% relative compaction by dumping it, in terms of relative density, a number along the lines of 30% might be more likely.  Now, relative density measurement with a fine-grained soil is not really possible because the "loose" state measurement is not possible.  With a coarse-grained (sandy soil) relative density measurements would be feasible.

My next thought and reminder is that a 1% change in relative compaction really represents a large change in stiffness or compactness.  I would believe it if an 85% to 92% (relative compaction) fine-grained soil is rather soft, or a coarse-grained soil in the same state of compactness is loose and visibly compressible under foot traffic.

breezeear;

What is your wet density, Moisture content,% Maximum proctor with optimum mc-we need these numbers to analyze your output.
Inintially, my initial guess is that you may be using bulk density divided by maximum density for this high output.

But, why are you testing compaction of organic soils? Organic soils should be removed and replaced with structural fill.  Also have you calibrated the gauge for that site. You mention that it is easy to push the rod without using a hammer. May be you're smearing the rod exterior and therefore affecting the count readings. Are you testing close to an elevation drop of more than 18"(like a shallow trench)?

MRM has pointed out a good distinction between relaive density and relative compaction. Since this is frequently occuring to you, it is best you share all the steps you take for the test, all the outputs and the lab results.  With this new information we'll all crack at the funny output numbers and hopefully put an end to your concern .  

Thank for the answers. First let me describe this project in a little bit more detail. This is a pretty big fill area for structures (approx 2000 ft x 2000 ft) close to a port. I don’t know for what reason the contractors was allowed to use the dredging material ( MH or ML, very fine material) from the bottom of the sea as part of the fill material (they mixed it with imported silty sands). But the mixing was of very bad quality. So the dredging material area caused serious pumping and settlements. Ruts can be observed as deep as 5” to 30” almost half area of the site.

The contractor hired one geotechnical firm to do the relative compaction control. Their NG and Sand Cone results showed every layer passed the 90% required minimum compaction. But why the pumping happened?

More than 20 test pits (typically 4 to 6 feet deep, 5 feet wide and 15 feet long) were excavated to do a checking. It is obvious the mixing was bad from the exposed pits. The Silty Sand areas look ok (pretty hard and Relative >90%). But the dredging matl areas are so soft. That’s the area I am talking about to expect 40% (I came up this unreasonable number just because I think it is so soft).

My NG test showed a relative range from 85 to 92%. They are maybe reasonable numbers according to BigH’s comments:” you get something like 80 to 85% "compaction" just be dumping it out of the truck”. My previous expecting 40% is too low.

Regarding to fndn’s comments about the procedure: When I did the NG test, first I did standard calibration in EACH Individual pit before I choose a “trench mode” then ran the test (I use the CPN MC-3 NG). I ran the test right below the spot where I did the calibration. (fndn, I didn’t push the test rod into the soil. I mean I used a probe, not the rod, sorry).

Regarding MRM’s comment :” I would believe it if an 85% to 92% (relative compaction) fine-grained soil is rather soft, or a coarse-grained soil in the same state of compactness is loose and visibly compressible under foot traffic.” This prompted me a question: for fine-grained soils, like in this project, even the test relative compaction passed 90% (like the other Geo firm did), but soil is still pretty soft, then the above fill layers may have pumping problem. Is what I saying is true?

The DRY density on my NG ranges approx. from 102 to 120 pcf and moisture from 14-21%. Relative compaction from 85 to 92%.  An optimum of 13% and a maximum DRY density of 120 pcf are ASSUMED for this kind of materials based on our experience I believe a dry density of 120 pcf is a reasonable number for the dredging organic or silty material. (Budget won’t allow us to run more lab tests other than soil moisture. We think NG can obtain an accurate soil density but it may not be true for moisture. Our lab soil moisture are generally 2-5% lower than the value obtained from NG. This is reasonable.)

Conclusion: I judge the density by the NG is ok (though relative compaction passed 90%). The problem of this project is that the soil moisture is on the wet optimum side. So, even the density is ok (passed 90% percent), but the soil is still very soft because of the high moisture content. This maybe the reason to cause pumping? Please comment.

Thank you for the additional information. When the fill placement began, if the Ground Water level was near the exposed surface or there was pumping with the initial lift, it is hard to stop it from following it upwards as the fill placement progresses.  The dredgeed material during grading is unstable because of the small window of moisture levels that it has to be placed in-it is either too wet or powder dry. If you had chunks white powdered soft rocks or clay lumps then it is even more unstable.  If the grader's equipment was not wide-track or low pressure equipment then that also hinders a bit.  

There is a problem with these materials as bearing materials, because they are loose, usually friction angles are less than 30 degrees and they are not uniform in vertical depth and in horizontal direction. So my approach would be as follows:

If the existing grade is lower than 3B from the bottom of the footings, I would place geogrids every 18" and bring another backfill materials that is either Sand & Gravel or silty clays.

If however you're too close to the proposed bottom of footing elevation, then either use piles or remove to depth of 3B, then place 12" of rock and then build it with suitable fill to the subgrade.  

As you describe these soils, they are prone to large settlements for structures, pavements and utilities.

One sign of hope is that should the earthwork grading finish as is, and the site is undeveloped say for another year, you could get some tightning. If this is your scenario, you may then take new borings, direct shear tests and do geotechnical analysis and reports to come up with allowable bearing pressures and shallow foundation design parameters.

ML dredgings, espically if higly organic are not suitable for fills. Even at 92% procter (how you got it there is a piece of work by itself) the material is an inherently weak material and probably does not have great shear strength at 100% procter. Couple that with the fact that these soils drain slowly and te organics degrade over time, I would venture that the site is a mess. Rather than fix it, I would suggest piles.

Just a note if this hasn't already been covered.  Compaction tests don't range from 0% to 100%.  0% would mean there was no soil!  Only air.  I have found it useful in the past to test the moisture of stockpiled material, that is completely loose material- no compaction, and I ususally get between 70-80%.  That's just how the ball rolls.  Just to let you know.  

If mixing is a problem, you are very likely to have at least three types of soils, which are silty sand, seabed fines and the mixtures  of these two materials. I think you have to use three sets of proctor valus, either assumed according to experience or tested to assess your compaction control.