Earthwork Calculation Discrepancy 1

[yrds3]

We recently started a fairly large earthwork project and are getting volume differences (3%) with the owner. They are using AutoCAD and LDD.

My questions are:

1. What are normal volume discrepancies? And how accurate is LDD compared to Agtek and Carlson?
2. I would like to keep good relations with the client but want to support my case. Any suggestions besides hiring a 3rd party?

Many thanks,

John

 

[cvg]

3% doesn't seem particularly high, especially when you consider that the design survey could be off by more than that. Also, the amount of shrink or swell or ground compaction could easily exceed 3%.

 

[LCruiser]

What did you figure for stripping during clearing and grubbing?

 

[Oson]

The field data should be taken together by both imterested parties, one set before you excavate and then again after excavation is finish, once both of you have the same data, and both of you agree what method to use, (the end average areas method is OK) there should not be any discrepancies in the end results.

 

[cvg]

"...there should not be any discrepancies in the end results..."

I disagree, there could be discrepancies depending upon how the existing ground TIN is created. If breaklines are not created at tops and toes of slopes, at inverts of ditches or channels or tops and bottoms of walls and steep slopes, the existing ground terrain model could be different when created by two different methods by different people.

Still, 3% deviation seems quite acceptable for earthwork quantity calculations. For a large project, you shouldn't expect better accuracy than that. I have seen contracts that do not allow any change in contract pay quantities for a deviation less than 10%.

 

[LCruiser]

On breaklines -

We had a roadway project where the breaklines were aligned to the hinge point (top of slope) of a roadway, but not to the catch point (bottom of slope). As a result, the contours of the engineered slopes looked really strange - a dead giveaway. The project was surveyed with lidar on 3m centers. Unfortunately, every time the laser beam hit a stick or something else protruding above the ground, it created a very short pyramid of air that the processing thought was dirt. Between the two errors, we were short 8,000 cy per mile. The project was over a hundred miles long.

 

[DRC1]

First I agree that anything under 5% isn't worth arging about. The calculations are only an estimate of the quantity, not a precise acounting. Check to make sure shrink swell topsoil and benching were all handled in the same way.
If so, I would just split the differences half and half .

 

[BigH]

While 3%, 5% even 10% are normal in estimates, make sure that this doesn't happen in measurements for payment when the project starts! Joint surveys by contractor and Engineer along with agreed on method of calculation of areas/volumes given in the specs.

 

[Beantown89]

I just performed an earthwork calculations using Carlson SurvCAD on a small site less than an acre site. I also used the average end area method. The SurvCAD amount was 6% greater than the average end area method.

 

[yrds3]

That's interesting Beantown. I'll try it with Agtek and check the variance.

I know there are three common methods of calcuation. I wonder which one is the most accurate?

Even 3% will make a difference of several hundered thousand dollars on this project ($4 dirt).

I'll let you know how it come in.

 

[yrds3]

That's interesting Beantown. I'll try it with Agtek and check the variance.

I know there are three common methods of calcuation. I wonder which one is the most accurate?

Even 3% will make a difference of several hundered thousand dollars on this project ($4 dirt).

I'll let you know how it comes in.

 

[TerryScan]

I find it interesting that people are just mentioning software packages vs. methods. For example, I know there are quite a few ways to calculate earthworks in Carlson's Survcadd. One can compare TINs...other methods use grids or interpolate grids from the TINs. When possible, I prefer to use TIN - TIN comparisons.