All MEP disciplines in ONE single model???

[Des]

I’ve recently started at a new firm and they are about to embark on one of their largest projects to date (circa $2 Billion) and I would like to try and get it right from the start but they want to do something really strange and I am trying to convince them otherwise but keep getting rejected because they are in the frame of mind that, “That is the way we have always done it”. Hence I am seeking some sage advice from all of you.

So, this project will have two towers (one residential and the other a hotel) with a 6 storey basement that connected to the city’s railway system. Now, Arup is a multi-disciplinary office and we will be doing all the Mechanical, Hydraulics, Fire and Electrical for this project.

They want to create a single model that contains all of our disciplines. I am arguing for them to split the models into their respective disciplines and link them together. They say that it is easier for the people to work in one model because they are trying to encourage their staff to move other disciplines equipment at will, if need be. They say their staff shouldn’t think of themselves as only Mech or only Hydraulics, they want everyone to be MEP. I don’t have a problem with this but I do disagree with them all being in the one model. They are still all in the same office and they all will have access to all the linked models anyway.

We are currently working on another fairly large project where they have got all the their disciplines in a single model and it is huge (500Mb) and really cumbersome to work with, lagging refresh rates and really long synchronise times. I fear that the new model is going to get worse and eventually so big that we are invariably going to have to split it into three parts anyway (Tower 1, Tower 2 and Basement).

What do you guys think about this and what do you think is the correct way to this?

Thanks,
Des

[Twiceroadsfool]

Well, there is the fact that the project is very very large, and then there is the extreme rhetoric you are using to describe their workflows. Its not as black and white as you make it out to be, so a few clarifications:

1. Its not uncommon for MEP engineers to all work in one model. I tend to give my MEP consultants the choice. Its up to them.
2. 500MB is not a large file. Not by todays standards. Especially since a few years ago, something seems to have changed in the "nuts and bolts" of Revit files that seems to make the files a bit larger, but they still perform fine.
3. I dont think its as simple as "this is always how it gets done." It doesn't have to always be one way or the other REGARDLESS of how much they want the disciplines to work together. But that means it doesnt ALWAYS have to be separate models either.

You mention two towers, and a podium, and you mention 2 billion dollars. I WILL say that 2 billion is a LOT of project, but there are still other factors i would be looking in to. How is the documentation getting packaged? Are the towers and the podium together? Or are there multiple sets? If so, that might affect how i would break up (if at all) the model. How many staff members to be involved? You get past a number like 10-15, and you start to have some spinny wheel effects in between every move, while people dance around each other.

But on the other hand, the hotel and the residential: Are you/they/everyone putting in every unit? And ever domicile? I would be if i was the architect on the job, BUT if the MEP Engineers arent, then there is no need to get worked up about the model, because half of it will be empty. Itll be risers and mains, and thats it. If they ARE populating the entire thing, KUDOS to them.

All in all, i think we need more information about the project, honestly. Especially if a lot of that 2 billion dollar number is infrastructure and underground costs at a railway station. A simple 2 tower residential/hospitality building? I could see it in one file, sure.

[Des]

Thanks Aaron,

I know the details are sketchy but I too don't have all the information yet. I just want to find the best way, but as you mentioned, there are going to be numerous influencing factors.

As for the project, the two towers are next to each other and the parking lot joins them together below ground.

If you do go the all-in-one model way how do you effectively deal with things like clash detection, as you can't then do it in the traditional way, using Navisworks, whereby you can create search sets etc. for each discipline.

Also, how do you issue models to individual contractors with all other disciplines in the model, most of which they will not need to see?

I also don't know yet how the documentation will be packaged, I think at this stage it will all reside within the project file, and there will probably be hundreds of sheets and countless views for those sheets.

At this stage I estimate that there will be around 20 people, from all disciplines combined, accessing the model.

Do you maybe have a checklist or something similar that will give me an idea as to how you would deal with something like this in your personal/professional capacity? A list of things to include or to be thought about in the decision making process?

I now realise that I have also only been thinking along the lines of what I have done in the past, using Revit in the "purist" sense but I am certainly open to change, to be honest I just never really considered putting everything into one model.

All the best,
Des

[Twiceroadsfool]

Why can't you use Navisworks in the traditional sense if it's in one file? That's just completely untrue, in every facet.

Exporting things to different contractors and subs? Not an issue at all.

It's also not more urist: or anything, to have them all separate. It's just different. That's it.

When I asked how it's being documented, I meant one permit package/set of drawings, or multiple.

The real issue (honestly) is that it sounds like you've made your mind up about what's right and what's wrong. There are a lot of questions to ask, things to consider, options to look at.

But it's not even close to accurate to say that putting MEP all in one model is rare, strange, weird, or outlandish. :shrug:

Sent from my LON-L29 using Tapatalk

[Des]

Yeah,

Maybe now was not such a good time to post my question, LOL. I mean, so soon after your confrontation yesterday with that other guy who hated Revit, as I am sensing hostility in your response.

I, on the other hand, LOVE Revit and all it is capable of and I am constantly trying to find better ways to do things, both with the software and within myself.

I absolutely have not made up my mind about what is right or wrong, I don't know how you deduced that from my post? I am open to suggestions and greet change with open arms but if I have been doing things in a certain way up to now and never had the opportunity to do it any different and now someone asks me to do it differently, by the very definition of that, the request will be rare, strange, weird or outlandish - to me (ones perception is ones reality). That is why I posted my question here to see what everyone else thinks or does.

You have given me a few things to think about though and I really do appreciate that.

Remember, not everyone has your level of experience and please don't post when you're mad

You're still awesome...

Thanks,
Des

[elton williams]

I'm regularly working on 500-1200 unit resi developments and since moving to v2017 I'm keeping models to a minimum. Current one is 2 x 60 - 70 level towers on 5 basements. Just started DD and its all in one model (+1 linked model for cad). Still orbits in 3D great with all worksets on, just need to show your staff how it works and how to not break groups. I set the starting view to a list of worksets/groups which people can refer to if they get lost.
I don't know how a $2,000,000,000 (just wanted to see all those 0's) MEP project would perform all in one model but our arch files are easily 500mb. Best thing I ever did was keep cad and too many links out of it.
A few years ago I worked on a 7 storey/ 63 unit project that had 63 links in Revit 13 and it was a b****. Still working on a 20 storey tower with around 70 links in Revit 15 and it's also a b*****.
As Aaron said there's been a LOT of speed improvements over the last few years.

[josephpeel]

I have kind of the opposite experience, I have always worked with a single model for MEP in my company and am now trying to convince people to split the model by discipline.

There are pros and cons for both options though.
If you are the only one working in a model, for a large number of disciplines, (Which I often am) it is great to be able to just change anything and coordinate on the fly. However I think the idea that this will work for more than a close team of max 3 people is crazy. You will just get "who moved what first without agreeing it" discussions and in one place the piping guy will have solved the coordination issues and in another the drainage guy etc.. so you will have a very inconsistent model. I think having separate models forces people to talk to each other and solve the issues properly, and also makes people more responsible for their own part of the work.

There are some things that could be a problem though with separate MEP models;

Connections between systems. Does a condensate drain for a fan coil go in the HVAC or Drainage model? How do you manage changes if the FCU is moved? To avoid a lot of manual checking of one discipline with another Im considering putting shared connectors in equipment that can be copy-monitored..

Spaces. If you use spaces or rooms a lot to show what objects are present in a room or to check against design data in the room/space, then this is only possible in the same model. So you will need to also have a master 'Spaces' model and somehow keep all changes consistent in your other MEP models.

Sheets. You can either let everyone make their own sheets in their own model (Then you have no central sheet list) or you can link everything into a sheet model and make the drawings there (So you are always working with linked models, so less direct control over the appearance of things).

Managing changes. If someone changes settings in one model, giving different appearance or behavior to the same items in another model, how do you mange this? I would say have a good template to start with so noone should need to change things, but it still happens.

I personally think separate models for disciplines, or even different purposes within a discipline, is the right direction to take but it requires some extra management to keep everything consistent and working.

[josephpeel]

... so big that we are invariably going to have to split it into three parts anyway (Tower 1, Tower 2 and Basement).

My advice on that would be DO NOT split the project on the basis of building areas, unless they are really completely separate projects in themselves.

Splitting by building, with all the disciplines in each model, means very little saving in size and complexity because you still need all the same stuff in each model. This also means you have to manage consistency for all disciplines in all models. Splitting by discipline means you only need the families and systems for that discipline, which means bigger savings in file size (Number of families has more effect on size than the number of instances of each). I did a large project that was divided by zone and not discipline and it was a nightmare to coordinate. (Even the Level heights, and pipe connections from one building to another ended up different between models !!!)

[Necro99]

We always put all MEP disciplines into a single model. The biggest problem with breaking it up into separate models would be losing connectivity. Our mechanical equipment families have "real" electrical connectors that use a parametric formula to drive the Apparent Load. This saves time for Electrical because all they have to do is power it up, they don't have to fuss around with individual junction box families into which they've hand-entered the electrical loads.

That's we've always done it. If I were told to make separate models I wouldn't know where to begin.

[Twiceroadsfool]

Sheets. You can either let everyone make their own sheets in their own model (Then you have no central sheet list) or you can link everything into a sheet model and make the drawings there (So you are always working with linked models, so less direct control over the appearance of things).

Just to clarify: This isnt really accurate. Everyone can make their own sheets, and all the sheets can still show up in a single drawing list (or in all drawing lists). Its NOT required that all the sheets be in one file to schedule them together, or no architect would ever be able to create a sheet list!