By: Ian Molloy, Senior Product Manager, Building Performance Analysis
Revit 2014 Update Release 1 (UR1) is now available for download from the updates section of Autodesk.com or from within Revit via live update. Anyone familiar with the annual Autodesk product cycle knows that while these URs can’t provide any new functionality in the form of user interface, data model or API changes they can contain updates to components that are independent of those things and generally provide improvements to performance and stability.
For the new Energy Analysis using Building elements feature in Revit 2014 the component that automatically creates the Energy Analytical Model (EAM) is independent and so we’ve been able to make certain improvements in UR1 that make it easier and faster to create a more precise EAM than before.
As I aim to illustrate here and I hope you’ll soon discover for yourself this update makes a big difference to the ease of use and precision of the resulting EAM for most Revit models. In summary the main aspects of the update are:
- ‘Sandwiched’ Revit elements – In Revit 2014 when two or more Revit elements were directly or closely adjacent and parallel i.e. ‘sandwiched’ (see later) you had to disable the room bounding property on all but one of those elements in order for the EAM to create representative analytical surfaces otherwise they would be omitted. This wasn’t ideal but now with UR1 this condition is automatically recognized and the analytical surfaces are created without you having to do anything. Note this update only resolves the geometry aspect of sandwiched elements and does not yet resolve sandwiched material layer thermal properties (if you are using them in Revit).
- Analytical Surface Precision – One notable product of our translation process is that the resulting analytical surfaces do not always extend to their full extents and converge thereby leaving small ‘gaps’ between the surfaces. Another characteristic you can often see are ‘rough’ analytical surface edges. Neither of these characteristics causes any problems to the energy simulation other than increasing simulation processing time but of course that’s one of the things the cloud is great for. Importantly these don’t allow air or heat to ‘leak’ or anything like that. In fact the only issue is that it can underestimate the surface heat transfer area slightly. Then again by adjusting the algorithms resolution settings to suit the model (see more about this later) these can typically be reduced to a level that has negligible impact on the Energy Analysis. In summary UR1 essentially provides more precise analytical surfaces with less gaps and rough edges thereby providing a more accurate representation of surface area for heat transfer and thermal comfort simulation.
- Faster processing – The time it takes to process the EAM when sending to Green Building Studio for simulation is now about 50% less than before thereby returning control to you in Revit sooner and allowing you to keep working. You may note the overall time to return simulation results is about the same but this is because GBS is now also processing ~40 additional simulations to provide the new Potential Energy Savings chart plus a means of viewing the EAM in Autodesk Design Review from Results & Compare. That’s pretty amazing when you think about it and it’s all only going to get faster!
Example – Process & Precision
The first two of these improvements i.e. sandwiched elements and surface precision are a little difficult to explain in words alone and so I’ve provided an example here to illustrate these more clearly.
The first obvious question – why this particular model? Well, while every Revit model is different and the precision of resulting EAMs will vary a bit depending on certain element configurations and settings I chose this particular model for a few specific reasons:
- It contains a set of sandwiched elements in the roof. This alone is not uncommon, which is why this improvement is so valuable but it’s easy to see in this model.
- It’s small but relatively detailed and therefore a reasonable test but you can also see the differences UR1 makes especially relating to the surface precision more clearly.
- I have a set of precise measures for things like floor area, volume and different surface types for this model to compare the automatically created EAM against.
While my main goal here is to illustrate the level of precision that can be achieved it’s also important to explain the steps I took (and didn’t have to take) to get there because apart from the comparison of Revit 2014 to UR1 comparison you should find it represents a good practice to follow to understand how the feature works and get the best possible EAM with the minimum effort.
The first and most important thing to note is that I did absolutely nothing to prepare this model for analysis i.e. no ‘cleaning’ or simplification. I started by simply running an initial simulation with Revit 2014 using the default ‘resolution settings’. These are basically just two parameters in Revit’s Energy Settings dialog – one that controls analytical space finding and the other that controls analytical surface finding. You can learn more about these in Revit’s wikihelp topic on model validation but I’ll be posting more here in the future on the role these parameters play and how to get the most form this feature. That said if you play with a few models yourself you will see it’s really not that hard. I’d recommend starting with some smaller Revit models and following the steps outlined here so you can learn fast. Alternatively, just ask for help. I’d be very keen to see the kinds of Revit models you are trying and after all that’s what we are here for!
This first run with Revit 2014 showed (see first EAM in the series below) that all spaces were found, meaning the envelope is enclosed (it doesn’t have to be perfect) but the sandwiched roof elements meant those analytical surfaces were omitted. As mentioned earlier this can be corrected in Revit 2014 by disabling all but one of the sandwiched elements room bounding property but that is less than ideal especially on larger models.
Next I ran a simulation using UR1 again using the default resolution settings. This provided a direct comparison of the improvement from Revit 2014 to UR1 (see second EAM in series below) which showed that the roof surfaces are automatically found and the surfaces are generally tighter. That said this EAM still showed some notable gaps which is where the resolution settings come in.
As I said earlier every Revit model is different and the resolution defaults simply provide a good balance of processing time and precision for most Revit models however greater precision can often be achieved by adjusting these values, usually downwards.
That’s why my next step was simply to reduce both these values down just a bit, run another simulation, check the EAM and if gaps still exist but all spaces are found then reduce the resolution values again until ‘balanced’ i.e. minimum gaps without missing any spaces in the shortest time. In this case it took me just two steps down from the defaults to the minimum settings to get an EAM that still had some gaps but they were very small and would have absolutely negligible effect on the Energy Analysis (see third EAM in the series below).
It’s worth noting; one temptation with settings like these is simply just to crank them down to their lowest setting but you should generally try to avoid that for a number of reasons i.e. processing time increases rapidly, spaces can be missed and it’s not always necessary.
This whole process took me about 30 minutes in total but the majority of that was processing time.
It’s worth noting that the views of each EAM I’m showing here were created from the new feature in Results & Compare that allows you to open the EAM in Autodesk Design Review.
So based on a visual assessment like those above can see when things ‘look right enough’ but just how right is it? This is where we can use those precise measures I mentioned earlier for this model to compare against. The table below provides a summary of the relative EAM precision for each of its key parts:
* Values relate to before and after applying the workaround for handling sandwiched elements with Revit 2014
You can clearly see from this comparison that UR1 makes a valuable difference and is very close in every respect to the real measures. This is especially true when you consider the average accuracy of manually authored EAMs in standalone analysis tools is approximately +/-10% and while that tends to be more a function of authoring practices done for purposes of efficiency than tool limitations it’s a valid reflection of current industry standards and processes.
No doubt this is a great result for this model but it’s worth noting that this level of EAM precision isn’t always possible or even necessary, especially in the earlier stages of the design process when Energy Analysis is especially important. Also this particular comparison isn’t something you need to do every model because it can be somewhat difficult to do in practice. I’ve simply provided this data here as evidence of the features effectiveness. The incremental reductions in resolution settings are important to consider though.
Of course this all shows why it’s still important to check and refine your EAM settings but that typically doesn’t take that long and importantly doesn’t involve cleaning or rebuilding the model. In fact more often than not I have found it actually helps you identify parts of the model where elements are missing (because spaces get omitted) which only assists in creating a better more complete BIM.
To close, I’ll soon be posting more on this entire topic of Building Information Modeling to Building Energy Modeling theory, tools and workflows but in the meantime I hope you find this update as interesting and valuable as I have found them a privilege to develop with a team that is continuing to work extremely hard on achieving the formerly impossible. We always need your input though especially to understand more about the way you work in Revit and want to fit analysis into your process so if you haven’t already please go get UR1, give it a try and let us know if you need help with your models – we’re here to help!