Cost Analysis

[nsinha73]

Can Revit Architecture be used as a cost analysis tool. Or do you use a separate software?

[Calvn_Swing]

In theory, it can be used as a cost analysis tool... In theory! It would work if you have met the following assumptions:

1. You have populated the Cost field for every type in Revit with accurate, up to date information

2. You've modeled every single element that needs a cost

Given those two monumental and unrealistic assumptions, you would be good to go. For the practical person, you know that neither of those two are possible. You won't model everything, and you can't keep that information up to date without a ridiculous amount of time devoted to it.

So, in general, I'd look to third party softwares. Our firm uses Innovaya's Visual Estimating, though our in house Estimating department uses it, not our Architects. Innovaya also has a product called Design Estimating which is probably what you're looking for. Both products link to a cost database, and can be linked to a managed database like RS Means so that you don't have to do that yourself. Then again, if your company has an estimating department they may already have a database and such in house. In which case, all of Innovaya's software links to Timberline. If you don't already use Timberline, you might look around for other options. I know that US Cost has something similar that links to their estimating packages.

Good luck!

[twiceroadsfool]

I will agree that it is a monumental amount of work to get it to work right, but i personally do not believe that you need to model everything that needs a cost, so much as have everything that needs a cost accounted for. :)

Its been my (very limited) experience that a great deal of cost is not only in the materials itself, but in how different items on the building get put on the building. For instance, would it really be beneficial to model in every stud in a wood framed house? Maybe, maybe not. Is it necessary for a cost estimation? Maybe, maybe not. I dont think so. I dont even think it has to be a function of the walls themselves, but maybe a function of some unit X, where X is a factor of both time and money. Maybe the wall has an X, and maybe each window has an X, relating to headers, sills, double studs, etc...

Very time intensive? Abso-friggin-lutely. Doable with a Revit Model? Definetely...

[Calvn_Swing]

"I will agree that it is a monumental amount of work to get it to work right, but i personally do not believe that you need to model everything that needs a cost, so much as have everything that needs a cost accounted for."

So in other words you do need to model everything that needs a cost? Aside from semantics, to take your example, while you wouldn't need to model every stud, you would need to model every wall, and include in your costs of said wall (per linear foot usually) of said studs, and unsaid sheetrock, nails/screws, joint tape, joint compound, insulation, and of course let's not forget labor! Don't take my comment quite so literally Aaron, I merely meant that you can't use Revit to cost something that you don't have modeled. This doesn't mean that you can't roll the costs for nuts and bolts into a single object that IS modeled. You do have to be careful doing this as these "assemblies" of costs (if they're constructed wrong) can cause a lot of problems.

That being said, my chief concern was never modeling enough to get a cost within 10% of a budget. I was much more worried about the upkeep of the so called "cost database" if it was stored in Revit content. Or, in trying (without some third party API plugin) to input costs from a good database into Revit for each project.

In all cases, yes it can be done. But, in all cases there are already programs that do it much better than Revit can natively, and tend to make the process much more streamlines and error resistant. (Not to mention more BIMish).

I think an even more important question is at what point are you making this cost analysis? Is it during project planning, schematic design, design development, or construction documentation. Alternatively, is this a conceptual estimate, a progress estimate, or a GMP? Or (like with us) is the answer all of the above.

Depending on the answer there are several solutions available. Heck, my company even went so far as to write our own stand alone application that does conceptual estimating and modeling targeted towards the first steps in project planning and conceptual design. If you're interested in it, PM me and I'll send you a link to the software (which we're now selling at a decent pace). If you're looking for something to get a GMP, then Innovaya's Visual Estimating or US Cost's competing solution are for you. If your answer is all of the above, then you're probably looking at Innovaya's Design Estimating (though it is limited in the early phases of a project by what information is available in the model.) Unfortunately, there isn't a one stop solution for estimating throughout a project's life cycle as the fundamentals are so different between a conceptual and a GMP estimate. Probably your best bet in this case is a combination of software. Something like our program followed by something like Innovaya's program all working with a BIM software like Revit.

Anyway, that's it for tonight...

[twiceroadsfool]

I see... I didnt mean to take what you had written so literally, i just meant that (depending on how specific of a cost, and as you mentioned, at what point in the project such a cost was being calculated), there are many ways to try to accomplish it in Revit... Some very simple, some not so much.

We could go nuts embedding information in to all of a Project Model, but we could also hash out an early estimation based on things like square footage of floor area, room sizes, or to get more specific... Yeah, a count and linear foot measurement of walls, doors, windows, etc.

I know theres a good deal of stuff people tend to not model, but i really wonder how high the percentage is? Obviously its tougher if there are things that arent in your work scope and therefore arent modeled, but i guess thats another issue alltogether...

[Calvn_Swing]

Well,

Our firm is "integrated" so I've been working in lock step with our estimating department the last few months to get our content and their cost assemblies to match up, and you'd be surprised how much we don't model that they cost. Given, this is a GMP estimate I'm talking about, not a conceptual estimate. We use our own software tool to do conceptual estimating and it uses very large assemblies of rooms and floor area and the like to get accurate costs. But even then the assemblies include things like how many screws per square foot and how much tape per square foot. Waste factors too. It's pretty intense.

From talking to them, they claim about 40% of what they cost is present in CAD drawings. On the up side, we're getting close to 60% in our Revit documentation, so that's an improvement for them already. We are going to use that Visual Estimating software to quantify all the stuff we do have modeled, and cost all of the stuff that is "standard" fare for us. This alone should reduce the time on an estimate by 60% for our estimators. So, we're pretty pumped about using a Revit model to do cost analysis and cost estimating through a third party app. But, we looked at just using Revit, and let me tell you we all had nightmares about what would be involved.

In terms of things being in the scope, since we're usually the GC on the projects we estimate, "everything" is in our scope to some extent, so we (the architects) usually model things in Revit that aren't in our scope persay. We're also training our construction staff so that they can start modeling things in Revit that they need (think site access and crane placement) for their work both in the field and for estimating.

We're going to do a case study in the near future to analyze whether is it faster to model one of our third party architecture jobs (whether we get CAD or Revit documents) using our own content and then estimate that model vs. using our traditional process or (in the case of a Revit file) using Visual Estimating on someone else's content. That should be a very interesting experiment indeed!

[twiceroadsfool]

Well,

Our firm is "integrated" so I've been working in lock step with our estimating department the last few months to get our content and their cost assemblies to match up, and you'd be surprised how much we don't model that they cost. Given, this is a GMP estimate I'm talking about, not a conceptual estimate. We use our own software tool to do conceptual estimating and it uses very large assemblies of rooms and floor area and the like to get accurate costs. But even then the assemblies include things like how many screws per square foot and how much tape per square foot. Waste factors too. It's pretty intense.

From talking to them, they claim about 40% of what they cost is present in CAD drawings. On the up side, we're getting close to 60% in our Revit documentation, so that's an improvement for them already. We are going to use that Visual Estimating software to quantify all the stuff we do have modeled, and cost all of the stuff that is "standard" fare for us. This alone should reduce the time on an estimate by 60% for our estimators. So, we're pretty pumped about using a Revit model to do cost analysis and cost estimating through a third party app. But, we looked at just using Revit, and let me tell you we all had nightmares about what would be involved.

In terms of things being in the scope, since we're usually the GC on the projects we estimate, "everything" is in our scope to some extent, so we (the architects) usually model things in Revit that aren't in our scope persay. We're also training our construction staff so that they can start modeling things in Revit that they need (think site access and crane placement) for their work both in the field and for estimating.

We're going to do a case study in the near future to analyze whether is it faster to model one of our third party architecture jobs (whether we get CAD or Revit documents) using our own content and then estimate that model vs. using our traditional process or (in the case of a Revit file) using Visual Estimating on someone else's content. That should be a very interesting experiment indeed!

Sounds like a great experiment! Id love to hear the outcome of it...

[Calvn_Swing]

When we get to it, I'll be sure to post something about it on the forums...