I like the idea of testpits for this investigation as it would give you an indication of what is likely to occur during the actual work if it is to be open cut. I see nothing wrong with your strategy. You may wish to look at past info from road construction if available to tighten up your program, if you think there will be too many testpits.

Testpits can create poor areas and you should use as small as possible bucket. Take care in recompacting pits and ensure that they wouldnot cause distress to adjacent areas with infrastructure. I presume that you have looked into these issues. A combination of drill holes and testpits may berequired if constraints are present. Individual judgment is required re site etc.

Thank you very much, VAD. Your recommendations such as checking past construction information and using smallest bucket are very helpful. I am a little worry about a digger is not able to dig deep enough to check the underground condition. Any suggestion? Thanks.

a 20 foot deep trench will require a large trackhoe or trenching machine.  It can be done, but will take some time, effort and lots of $.  That is why investigations are typically done using many borings and a few pits.  I would recommend typical spacing for borings at about a 1/4 mile.  If you suspect areas with shallow bedrock, you might consider additional borings to try and define the limits.  You might also consider seismic refraction if you anticipate much rock excavation.

Thank you, cvg! It is a very valuable suggestion.

To go to 20 ft, the backhoe will have to dig itself down. This can be done using a Cat 235 or 245 hoe or equivalent.
I have used such hoes to testpit to 7 m. I still suggest a combination of testholes and testpit approach. If you want to do testholes then use a Texoma rig that give a large boreholefrom which observations can be readily made.

Some municipalities and county governments have minimum spacing of borings.  In Michigan it is typically 300 ft. spacing and 5 feet below the invert level of the pipe.