In construction, there are many methods to accomplish the same result. The main requirement is that the work is done according to the specification written for that particular project. When it comes to trench backfill, the specification usually allows the contractors several choices for filling the area over the pipe and under the pavement. But those of us working in the field long enough have probably come to prefer one over the other. After watching trench performance for about 20 years, I have come to prefer sand backfill compacted by some mechanical means.
Compact
The sand/compact method I prefer is described best in the Standard Specifications for Road and Bridge Construction in Illinois. This spec allows for material meeting the following requirements: a wet, coarse aggregate gradation such as CA 6, CA 10, and CA 18 or a moist, fine aggregate gradation such as FA 1, FA 2, FA 6, or FA 21. This material is then compacted using some mechanical means such as a compactor on an excavator or a "jumping jack" type compactor. My experience has been that, if properly compacted, the pavement over the trench has minimal chance for failure. The video below shows a person compacting around a pipe with a jumping jack.
Jet
In my earlier days, I inspected projects designed by other engineers. Some allowed what is known in our industry as "water jetting." This practice allows the backfill of the trench with specific material but no compaction. Instead the contractor returns to the trench at a later date and forces pressurized water into the trench using a long, metal device. On the projects this method was allowed, I noticed that within a couple years, the area of pavement over the trench would "dip" or fall creating a depression in the roadway.
My thoughts on water jetting are that it might work in the right soil conditions such as in sand or in a highly fissured bedrock. But most of my projects have involved the installation of pipe in heavy clay. Not necessarily the best soil type for conveying water out of a trench. So I envision the water introduced by jetting as filling a bathtub. Only after a few years does the water dissipate, reducing the volume of the trench, and causing a pavement failure. This is why I do not allow this method on my projects.
Fill
Another material used to backfill the trench is referred to as flowable fill. This is a cementious material with a low water/cement ratio that is delivered to the jobsite by a ready-mix truck. 
The material flows straight from the truck chute into the trench.
In my experience, flowable fill has performed well. The primary tip to remember, if choosing this method, is to make sure the pipe is covered first with an aggregate material. Allowing flowable fill to engulf the pipe can create problems for those needing to hand dig around the pipe in the future. The other challenge with using flowable fill is that the material does flow. So the contractor needs to block off sections of the excavation each day to allow the material to fully fill the excavated trench and remain out of the area left unexcavated for the next day. The contractor also needs to plan his backfill operations so they are synchronized with the delivery of the flowable fill.
Dump
The last method is one I have just started seeing used by contractors within the last five years. This involves the dumping of an open graded coarse aggregate such as CA7 or CA11 in the trench with no method of compaction. While I have seen a few specifications written by consulting engineers allowing this practice, I have not seen a government agency with this specification although there could be some out there. My concerns with this method have been that the open graded backfill has voids into which fine material from the soil can migrate. Of course this would happen over time and instead of the area over the trench failing, the area just outside the trench would fail due to a loss of material. And I have seen this failure occur on a privately funded project. However, I have talked with other engineers who have not witnessed any failures. I would agree that lining the trench with a fabric would allow this method to work.
Contractors seem to prefer this "dump" method because it requires no mechanical compaction and therefore eliminates the time spent compacting the trench. Their argument is that the stone compacts itself, and the trench will not fail. While I would still think the stone needs some compaction, I would agree that I have not seen trenches backfilled in this manner failing. Instead it is the area just outside that fails.
While I have, as an engineer, made my decision on the methods I feel comfortable specifying, I realize others have chosen to spec some of the others. I would be very interested in hearing the choices made by others, the reasons behind the choice, and their experiences. Feel free to leave a comment or take the quick poll on our main Website page!
