Revit 2010 MEP Worksets

[gvluisi]

We are gearing up to use MEP on a small project. First time. We will be using the M and the E but no P. I am deciding to use one model and enable worksets but not sure how to break the worksets up, obviously by discipline. What are some pros and cons doing it this way vs. a separate model for each discipline. Thanks in advance for your thoughts.

[RobertB]

Well, the pro is that you can utilize the cross-discipline connectors on the families. The con is that your team will need to be disciplined (groan) to make sure their stuff is going to the correct workset, although that is easy enough to correct when needed.

[kyle.bernhardt]

Why do you feel you need to break up the model by discipline? I know Worksets are used for a number of different things beyond their designed usage (Visibility Control primarily), but I'm interested in the reasons for this approach on your first project.

My recommendation (and the recommendation of the Revit Platform Technical Note), is to let Element Borrowing do the work for you.

I have seen BIM Managers create a workset for "Really Important" stuff like Levels, Grids, Linked Files, etc. They check that one out so others in the project can't modify those Elements.

You can create a logical Worksets structure if you'd like, I just want to make sure it's not being done to manage Element Ownership. That's not the best way to manage things.

Cheers,
Kyle B

[RobertB]

You can create a logical Worksets structure if you'd like, I just want to make sure it's not being done to manage Element Ownership. That's not the best way to manage things.Excellent reminder!

[gvluisi]

Like I said, First time using MEP. My thought was to limit linked models. The arch's the mech's and the elec's... so I link only the arch's model into our Eng. model and I only have to manage two models. There will probably be one guy doing the Mech work and another guy doing the Elec work. and the entire office watching to see of this bird is gonna fly.

[smcmillin]

Oh it will fly, but you will soon crash on an island wishing you were back home (Autocad) and then once you get off the island, for some reason you are compelled to go back to the island (Revit) and once you get there you want to go back home again :)

[kyle.bernhardt]

I agree with your logic in using one model file for MEP, instead of many. That wasn't in question.

Within that one model file, my previous point applies. You don't need to create Worksets to facilitate Element Ownership, just use Element Borrowing.

What is Element Borrowing you ask? Check out the Revit Performance Platform Technical Note (http://images.autodesk.com/adsk/files/revit_tech_note.pdf).

Cheers,
Kyle B

[gvluisi]

I enjoyed your comparison of Revit to the tv show "LOST" but I prefer the one where the guy wakes up with special powers (ACAD) but when he uses his special powers he ends up wiping out the human race only to rebuild it (REVIT)

[Misteracad]

Why do you feel you need to break up the model by discipline? I know Worksets are used for a number of different things beyond their designed usage (Visibility Control primarily), but I'm interested in the reasons for this approach on your first project.

My recommendation (and the recommendation of the Revit Platform Technical Note), is to let Element Borrowing do the work for you.

I have seen BIM Managers create a workset for "Really Important" stuff like Levels, Grids, Linked Files, etc. They check that one out so others in the project can't modify those Elements.

You can create a logical Worksets structure if you'd like, I just want to make sure it's not being done to manage Element Ownership. That's not the best way to manage things.
Kyle, I'm just curious if the above recommendation, or the documentation mentioned, is specific to which version of Revit is being used...2010 versus 2009 or 2008.

Reason being, when we first embarked on the Revit path using 2008, our VAR recommended separating all of our MEP disciplines into different models...Mechanical, Electrical, Plumbing and Fire Protection. The "pilot" project at hand was fairly large...a soccer stadium of all things...and was managed by an architect that had several Revit projects under their belt...but none that involved MEP collaboration :cry:. That is a story in itself, but what we ended up with was 4 individual MEP models with a LOT of complexity and no real continuity between disciplines. I questioned the approach more and more as I followed these threads, but it was too late to turn back once we got started. Plus we were concerned about hardware performance if we had gone the other way.

FYI the individual MEP models were worked on by different users in multiple satellite offices across the WAN utilizing multiple T1 lines and Riverbed. We supplied our users with high-end Vista 64-bit machines containing 4GB of RAM, but even then we had performance issues. If we had not broken the models up, I can only imagine things would have been worse performance-wise, yes?

I should also mention that we were fortunately...depending on how you look at it...able to convince the architect to upgrade to 2009 early in the project due to the MEP improvements we read about, and all the models were eventually upgraded across the board. That helped in some ways, but made matters worse in others. In the end, users and managers were less than optimistic about the true viability of BIM using Revit at this stage of the technology, and it was hard to argue with them.

Obviously I am aware that it is not accurate or fair to judge the technology after one, less-than-ideal project to work with, but people tend to make a lot of decisions based on first impressions unfortunately. Looking back now I wonder how things might have been different if we had kept the MEP model in one piece and used the workflow you describe. Any thoughts? Thanks in advance!

[mwiggins121466]

Tim,

In 2008 you would not have wanted all in one. You would have not been able to draw a duct or place a light due to performance. We have done about 20 models to date and we keep our M and E seperate.

[kyle.bernhardt]

Tim,
Sorry to hear that you had a less than stellar experience with that project. I'll try and address your questions.

The recommendation specifically applies to 2009 and 2010, as these products were used during our internal investigations, and the basis for the discussions we had with our internal folks and customers like you.

For Revit MEP 2008, the same recommendations could very well apply, it just wasn't the focus so I can't say for sure. However, it's safe to say that for a medium to large size project in Revit MEP 2008, it would be necessary to implement some type of a multiple project strategy to mitigate performance issues. We would have just recommended a multi-discipline approach than a per-discipline approach.

This is not to say, however, that it is a black and white situation, where one approach is "wrong". There are actually shades of gray in this discussion, and arguments for and against either option.

After discussions with customers, our Customer-Facing teams, and the Product Team, we felt that a multi-discipline strategy was most appropriate for most projects and most firms. We go into more detail on this subject in the Revit Performance Technical Note (http://images.autodesk.com/adsk/files/revit_tech_note.pdf).

At the end of the day, it's a choice by the Project Team, and we just try and equip you guys with the best information to make and informed decision. There are performance gains, workflow impacts, collaborations impacts, etc. that must be taken into account for each approach.

The underlying concern in most cases driving a multiple model strategy is Performance, and that is a detailed subject in such a Parametric BIM application like Revit.

The perception of slow performance can manifest itself to the user in many different operations in Revit. Each of those operations may have entirely different cause for a perceived performance issue. Thus, it's not appropriate to take a broad brush approach to mitigate performance.

Two of the big ones, View Performance and Model Manipulation Performance, are dealt with in the MEP section of the Technical Note. File Open/Save Performance is handled in the other sections.

If you go through that and have a specific question, I'll see if I can address it for you.

Cheers,
Kyle B

[ray.205827]

I am presently working a one week ontract for a local engineering firm. They have worksets by discipline already set up. Today, I had the piping workset open as did another user so we had a battle royale. No surprises there. Odd thing was that another user who had the plumbing workset active was also having difficulties with myself over element ownership and vice versa. Would I be correct in surmising that this would be less of an issue with no worksets. The central mep model is 130mb. How is it that I was showing up as the borrower of the plumbing workset in the worksets dialog in addition to various other worksets I had not touched?
At another place I was contracted at brefly, they had a three building project with a central file for each building and separate worksets in each building not only by discipline, but also by floor level. Using the split tool for all risers was a bit of a pain. So it seems that worksets are almost universally used and that resellers push the concept quite a bit.

[rrubert]

Like Kyle said, I don't think there is a "right" way to set up projects. It is very project specific. As performance gets better, more and more projects will be able to be in one model. To date, we have not split up models into discipline specific models, but we wished we had. The model performance technical note mentions that at a certain file size (160 MB), you should split up the models. I think that should be more like 80-100 MB for the cut off, at least in 2009. Revit 2010 has a lot more performance enhancements.

As for worksets, I don't think there's a right way either. It is job specific. However, I highly recommend element borrowing versus workset checking out wherever possible. As a minimum, I recommend one workset per discipline (assuming all disciplines are in one model), so that you can use workset visibility to control your views. If there are different wings or buildings that are served by separate systems, that situation could lead to one workset per wing/building. I think resellers as suggesting one workset per floor per discipline, and that has never worked well, at least for us. Another benefit to worksets, according to autodesk, is that you can turn them off and not have them load when you're opening a file, thus making the file load time quicker. We haven't really seen a benefit to this, but perhaps some people do.

[kyle.bernhardt]

Ray,
Before I address your specific questions, it seems there is a point I haven't been clear enough about.

*******************************
The existence of Worksets in a project is not a problem in itself. The problem arises when users begin making entire Worksets Editable (often called "checking Worksets out"), as a method of managing what Model Elements they can modify.

This was the original practice in Revit when Worksharing was implemented, but Element Borrowing has been subsequently implemented in later releases of Revit, and is the better way to manage Element ownership.
********************************

Making an entire Workset editable, to use an AutoCAD analogy, is like opening multiple DWGs at once, just in case you need to work on all of them. If another user wanted to use one of them, which you weren't working in, they'd have to wait for you to close it. Not the most productive workflow, which is why people don't do it.

Element Borrowing is akin to just opening the DWG you need to work on, leaving everything else able to be opening by other users. It minimizes your "Element Ownership Footprint" in the Revit Project by only making Editable Elements that you have actually requested ownership of.

Now to your questions:


Would I be correct in surmising that this would be less of an issue with no worksets.

Maybe. But that's not the point here. It's not the Worksets that are the likely problem here, it's the workflow being used. Worksets can exist to serve other purposes like Visibility Management (which they actually weren't design for), just don't use them as the mechanism of managing Element Ownership.


How is it that I was showing up as the borrower of the plumbing workset in the worksets dialog in addition to various other worksets I had not touched?

You may not have touched them directly, but changes you made to the model, and the Model Regeneration that accompanied those changes, resulted in you needing ownership of those Elements.

What version of Revit MEP are you running?


At another place I was contracted at brefly, they had a three building project with a central file for each building and separate worksets in each building not only by discipline, but also by floor level. Using the split tool for all risers was a bit of a pain. So it seems that worksets are almost universally used and that resellers push the concept quite a bit.

As I said earlier, Worksets can be used for many different purposes (many of them beyond the original intent of the feature). That in itself isn't wrong. It's when their usage impacts project productivity that it becomes a problem.

Good BIM Management and planning should help to mitigate the issues, and ensure a successful project.

Cheers,
Kyle B

[mjdanowski]

We have been using worksets for visibility as Kyle has alluded to a few times. However, as we move on to new projects I am considering just going back to category view settings for my electrical views.

Mechanical still uses them to distinguish between hydronic and plumbing piping, but electrically (in my opinion) they are not needed.

Category views makes it a lot easier to turn certain elements on for coordination without having to deal with the whole workset of a discipline.

[mwiggins121466]

When we started we used worksets by floor and system. We have found that you can control visibility with just systems. So for us it is power, lighting, and misc systems.

[kyle.bernhardt]

Mechanical still uses them to distinguish between hydronic and plumbing piping

Seems like a less than ideal solution to this problem. Why not just use View Filters?

Cheers,
Kyle B

[Misteracad]

Tim,
Sorry to hear that you had a less than stellar experience with that project. I'll try and address your questions.<snip>

After discussions with customers, our Customer-Facing teams, and the Product Team, we felt that a multi-discipline strategy was most appropriate for most projects and most firms. We go into more detail on this subject in the Revit Performance Technical Note (http://images.autodesk.com/adsk/files/revit_tech_note.pdf).<snip>

If you go through that and have a specific question, I'll see if I can address it for you.
Thanks for the detailed response Kyle, I will review the document and let you know if I still have questions going forward. Note I thought I had already read the white paper on performance, but perhaps I am thinking of one published last year for 2009? I'll be curious what I take away from this version that is different for 2010-based product use.

[rrubert]

Seems like a less than ideal solution to this problem. Why not just use View Filters?

Cheers,
Kyle B


What about your comment that, "Worksets can be used for many different purposes (many of them beyond the original intent of the feature). That in itself isn't wrong. It's when their usage impacts project productivity that it becomes a problem."

And me using worksets for visibility control is not impacting productivity. In fact, it is increasing productivity, as we don't check them out and use element borrowing (see my previous post).

View filters are good idea in theory, since I have different view filters for all the different piping systems. I could technically turn off the plumbing piping systems in the HVAC views. I will look into that on future projects.

However, I do not always add every plumbing element to plumbing systems or every piece of HVAC equipment to the HVAC systems. Doing so would take a lot of time. So, for now, it is quicker to make the worksets. Also, though we use worksets, we never check them out, so we still just use element borrowing (see my previous post). Another reason is that some duct accessories and mechanical equipment are cross-discipline.

I read the Hybrid MEP Design and Documentation White Paper from Autodesk today, and I feel that most of my "less than ideal" ways of doing things are part of the "hybrid" solutions given in this white paper. I do appreciate advice from you revit gurus, and I am always on a quest to find better ways of doing things, but I have to "leverage" the benefits of revit with the real time constraints given to me for a project.

[kyle.bernhardt]

View filters are good idea in theory, since I have different view filters for all the different piping systems. I could technically turn off the plumbing piping systems in the HVAC views. I will look into that on future projects.

However, I do not always add every plumbing element to plumbing systems or every piece of HVAC equipment to the HVAC systems. Doing so would take a lot of time. So, for now, it is quicker to make the worksets. Also, though we use worksets, we never check them out, so we still just use element borrowing (see my previous post). Another reason is that some duct accessories and mechanical equipment are cross-discipline.

My assumption was that you were creating Systems, so View Filters effectively deliver an automatic management of visibility where Worksets requires an additional management task.

Without Systems, Worksets is an effective way to manage this, I just wasn't thinking about it that way.

As for creating Systems...we don't need to go there. That is something that has been discussed ad nauseum here.

Cheers,
Kyle B

[mjdanowski]

My assumption was that you were creating Systems, so View Filters effectively deliver an automatic management of visibility where Worksets requires an additional management task.

Without Systems, Worksets is an effective way to manage this, I just wasn't thinking about it that way.

As for creating Systems...we don't need to go there. That is something that has been discussed ad nauseum here.

Cheers,
Kyle B

Even with systems though, it can get kind of time consuming making sure all the filters work, systems are complete, and intent is properly defined.
Then there is the issue with new users to Revit having a harder time with the concept of filtering compared to worksets.

Filters and systems are definitely an alternate approach that might be advantageous in some implementations, but worksets just seem to give us a lot less trouble on the piping end.