accuracy of volumes

[cldowney]

If you run all 4 methods of volumes. How accurate are those numbers? Of course if your surfaces are accurate?
Is there a number, for example 5% accuracy?
Thanks!!
Connie

[mjfarrell]

Connie,

Given that ones surfaces are accurate, and one properly applies
the known adjustment factors a margin of <than 5% is achievable.
Great question, care to share with us what has prompted you to ask it?
I think there is a bigger story here.....

[Jeff_M]

When I used to run my own Eathworks, I would be +/- 3% of actual dirt moved. I used the average of the Grid and Section methods. There is now a local guy that does most Earthwork calcs for the majority of the subdivision contractors. He uses AgTech and I've been told by a few of the estimators that his numbers are not as close as mine were.

And as you stated, it can only be as accurate as your surface data.

[cldowney]

It is a long story, but I'll try to give you the short version.
We are doing 32 landfills for Puerto Rico and we are getting volumes for waste.
There are several offices doing the calculations differently. We have a liner and a final cap for our landfill that both cover the whole site and are both 1 meter in thickness, but have different side slopes, of course. The ones doing the calculations are getting the same volumes for both of these two surfaces or layers. So the engineers are thinking that LDD is just not accurate. I think it is the procedure used by the user. I haven't had any problems.
I feel like I'm babbling, I hope this made alittle bit of sense.
Thanks for your help!!
Connie

[MHultgren]

It also depends on the spacing of the data when computing volumes. If your points in the TIN are 5 feet intervals, you will get a much better figure than if they are say 50 feet intervals. Run a grid volume of the same surface with different m and n numbers and you will see what I mean.

[pmahany]

Volumes. If the 1 meter is a vertical distance, then the slope don't matter. Of course. on a steep slope the distance perpendicular to the slope will be less by the function cosine of the zenith angle of the slope. So the question goes back to the definition of thickness and how the model is set up. Question, how complicated are your parcels that you are using for your volumes. The reason I ask is that for the stuff that I do in order to get results I will do the actual volume calculation in R14.


Paul Mahany

[mjfarrell]

The reason I ask is that for the stuff that I do in order to get results I will do the actual volume calculation in R14.


Paul Mahany

The data that you sent previously did not present any challenges to performing
volume calculations in Land Desktop 3, or 2004. I have not tried your data in
2005 as yet, however I have no doubt that the results would be as accurate and
complete as those from the other versions. The newer version will perform the
calculations without resorting to R14.

The difficulties you are encountering appear to be more process related than
programming failures.

[pmahany]

So, does methodology mean the way the equipment works, or the particular method of working? I seem to recall that you had to fidget with the parcels for sub volumes. I have been working in 2002 quite a bit, including setting up alignments and profiles and creating the basic 3D model that I use. Due to several factors, one being the way Grips have been dumbed down, and the sub volume issue. I find that I can be vastly more efficient doing those particular things in 14. EX. I got a landscape base yesterday at 2 pm. 11am today they are building it. Meandering walks, ADA, and it's a hill. Done, in 3D, in 14.

[mjfarrell]

So, does methodology mean the way the equipment works, or the particular method of working? I seem to recall that you had to fidget with the parcels for sub volumes. I have been working in 2002 quite a bit, including setting up alignments and profiles and creating the basic 3D model that I use. Due to several factors, one being the way Grips have been dumbed down, and the sub volume issue. I find that I can be vastly more efficient doing those particular things in 14. EX. I got a landscape base yesterday at 2 pm. 11am today they are building it. Meandering walks, ADA, and it's a hill. Done, in 3D, in 14.

I believe it is the tools(within Land) and the manner that they are employed
either the sequence, or manner that their outcome is combined.
It was not the parcels that I manipulated, it was some of the manner that
certain aspects of the grading design was created.

I too miss 'COLD" grips from the old days, however with good use of Snap Magnet,
and other improvements to the basic drafting tools one can function quite well without
them.

I truly believe that at issue here is that one should have the the same or better
results in design and production capability in 004/005 versions of Land.
There are a couple of caveats here, these gains are simply not achievable without
thorough training in the new tools. The training should be approach with an open
mind in regards to process and methods, as the new tools allow one to approach
the design concept to finished plans in a different manner. The training would also
need to address changes in the basic drafting functionality of AutoCAD.
The type of site design you speak of could clearly be done in the time indicated
in 3d in the new versions, and probably in under 4 hours.

[pmahany]

Well I seem to be too busy. I thought I had posted a reply and it seems to have gone awry. So I'll try again. Actually I did the 3D modeling in 2 hours using 14, the rest of the time was involved with plotting a one sheet plan and meeting with the Client and issuing that plan to the field.

ON ACCURACY

I have just had an experience that goes to the heart of the Accuracy Question. It involves a portion of "Big Project". I was just given a topo done by the grading contractor. It had been used 2 years ago by the grader to justify an ~110,000 yard import at the time. At that time I had said that the site, with 2 million bank yardage was essentially at balance. Today, there are several large piles of dirt out there. At the request of the Client, I received the 2 year old topo last week. I analyzed against the 3D model that I maintain. I used a 5' Grid on my Mass Grading Model. I predicted that the piles would total 80,000 yards. We topoed those piles after the prediction. The piles came in at 81,000 yards. The error of the prediction is 0.2% based on bank earthwork outstanding from 2 years ago. Now this could be a fluke, but this is not the first time this has happened. What does this mean? Well, one could say that accurate model drawing is at the heart of the methodology. I might suggest that K.I.S.S. as an operating principal is worth more than unnecessarily complicated and kludgey thinking. Especially in Software. For whatever reason the results I got 14 were unattainable with the same ease in 2000whatever. And last but not least, I can say that the method is validated by the results.