Workflow for Creating Closed Loop Hydronic System

[Iev60047]

I am using Revit 2019 MEP. I am trying to convince my department that it is worth our while to use Revit for piping pressure loss reports, but am having difficulty figuring out how to properly set-up and troubleshoot a closed loop hydronic piping system. I am currently testing a simple system with one pump, and two water source heat pumps (WSHP's). As I have not found a source that concisely lists the steps/workflow for setting up a simple system like this, I am posting here for documentation and to receive feedback.

Some things I have learned along the way:

1. Ensure that under MEP Setting, in Mechanical Settings that the checkbox for 'Enable analysis for closed loop hydronic piping networks' is checked. Otherwise your piping system pressure drops will not report.
2. Connectors for all mechanical equipment needs to have the flow-in and -out connectors linked to one another.
3. If certain pieces of mechanical equipment are 'driving' (setting) the system flow, like the WSHP's for my example, then: 
   1. Pumps should have their connectors' Flow Configuration set to Calculated (I think?).
   2. Pumps should have their connectors' Loss Method set to Not Defined (I think?).
   3. Pumps should have their 'Out' connector's System Classification set to Hydronic Supply (I think?).
   4. Pumps should have their 'In' connector's System Classification set to Hydronic Return (I think?).
   5. WSHP's should have one of their connectors' Flow Configuration set to Calculated and the other to Preset.
      1. You will need to manually enter your Water Flow into the Object Property back in the project, found in the Mechanical category of the object's properties.
   6. WSHP's should have both their connectors' Loss Method set to Specific Loss.
      1. You will need to enter your Specific Loss (Pressure Drop for piece of Equipment) into the Object Property back in the project, found in the Mechanical category of the object's properties.
   7. WSHP's should have their 'Out' connector's System Classification set to Hydronic Return (I think?).
   8. WSHP's should have their 'In' connector's System Classification set to Hydronic Supply (I think?).
4. Once all these variable are set, then you are ready to create your piping system(s). For this example you should create one (or two? I do not know) piping systems and generate a layout in order to calculate your piping pressure loss report.
   1. If you have already creating piping systems before this variables were set, deleted the piping systems associated with the equipment and piping, set the variables and attempt to set the piping systems again, you will run into difficulties.

I apologize in advance if this tutorial/workflow is available already at some other location on this forum, or in another forum, but I was unable to find the information on my own (I welcome any suggestions on resources that is concise and pertains to the same topic!). Anyways, does anyone have any suggestions, recommendations, and or additional information that would correct and/or complete this workflow? Thank you in advance.

[RobDraw]

One thing that sticks out to me is the pumps having supply and return. They are usually set to the same system supply or return. Only two places need both. The source and the end of the loop. If you are serving multiple units, it could be a balancing valve. If it's a single piece of equipment, it would have supply and return. 

Sharing the model for the sake of discussion would help getting everyone on the same page.

[Iev60047]

I am attaching two Revit Project Files and a few Pressure Drop Reports. The 'Piping_Systems_Test_Share_Old.rvt' is set in a way, where there is a separate supply and return system; the other .rvt file contains only one system that is a supply system. The Pressure_Test_Report.html is associated with 'Piping_Systems_Test_Share_Old.rvt' and the Pressure_TWS-2_Test_Report.html is associated with 'Piping_Systems_Test_Share.rvt.' 

Thanks in advance @RobDraw 

Pre-Post Edit: Html's converted to pdf's and placed into zip folder.

[RobDraw]

That's not a complete water system. You need another piece of equipment.

[Iev60047]

Not trying to waste your time here @RobDraw, but exactly what element do I need to add to make this system complete? I was under the impression that Revit could handle this system exactly as I modeled it, regardless if it is realistic and/or practical. Do I need a 2-way control valve for combined flow sections of pipe, or a 3-way control valve for water supplying WSHP's? Or something else completely? Thanks again.

[RobDraw]

Don't you need something to heat the water?

[Iev60047]

@RobDraw Is a cooling/heating source/sink necessary in Revit to properly calculate pressure drop in a piping system? For example, the boiler is categorized in its family as Mechanical Equipment, so I don't know how Revit differentiates it from the pump (which is also categorized in its family as Mechanical Equipment). Unless, it has something to do with OmniClass Number, because the standard Revit family doesn't include any parameter info that indicates it's delivering heat to the piping system.

I attached a boiler to the Revit project file with just one system currently defined and am sharing this file with you.

Edit: I altered the system, so that the boiler is now plumbed in series. I also changed all the connectors' types to global, except for the boiler, where I set the pipe out to Hydronic Supply and the pipe in to Hydronic Return. The system is still not working to calculate a pressure drop for the system.

Edit 2: From the Revit support page on Hydronic Piping Networks, boilers and chillers are optional in a hydronic piping network:

Hydronic piping networks are defined as having: 

• A single source equipment component, such as a boiler or chiller. This is optional.

[Iev60047]

@eric_grey I see that you have answered similar kinds of question in the past. Would you have any idea what alterations to the above workflow would be necessary to correctly set up a simple system like this? And is a cooling/heating source/sink necessary to establish the return and supply systems for a closed loop hydronic system? Shouldn't setting the connectors on the pump to either hydronic supply or hydronic return be sufficient for Revit to create two different systems?

Edit: Reading the support page on Hydraulic Separation, it seems that maybe I cannot calculate the piping system pressure losses in my system for two reasons:

1. There is no bypass after my boiler, bypassing the WSHP's - it is therefore impossible to isolate the primary loop (in my file it is the boiler and the pump) from the secondary loops (the two secondary loops would simply be one WSHP each) using hydraulic separation.
2. Furthermore even if I had the secondary loops set up, they do not each have a pump which seems to be a requirement.

I am also confused by the first sentence of the support page: Hydraulic separation provides flow and pressure loss calculations for piping loops. - Does this mean that hydraulic separation is necessary to calculate flow and pressure loss for piping loops? Seems ambiguous.

[Kevin.Lawson.PE]

This is the most important document with respect to flow propagation that you will find: https://thebuildingcoder.typepad.com/files/me204-3_connectors_in_revit_mep_content.pdf

It looks like you are getting your flow propagating already, which is half the battle.  The other half of the battle is getting all of the correct data into the model.  You need to use the correct piping material with the correct internal diameter and roughness set so pressure drop for straight lengths can be calculated properly, then you need to ensure all of the fittings you use have the correct K Coefficient Table set.  The K coefficient tables all come from this book: https://estore.pumps.org/Guidebooks/EngData.aspx they are all correct and work perfectly (unlike the duct fittings, but that's a different post).  

The final aspect which stops most people is that in order to get accurate pressure drop calculations, you would have to model all of the specialties that lead up to the unit (service valve, balance, valve, control valve, reducers, ect.)  essentially every unit would have to be modeled like something below.   

This may be excessively time consuming, you can work around adding the specialties and manually inputting the unique terminal unit pressure drop by just giving an allowance at each unit.  You can just say, I'm going to allow a 15' pressure drop at each terminal unit to account for the coil and all the piping specialties, then you can assign that in the family and you don't have to model all the specialties.  

All of this takes some time to set up and practice, but once you get it running it's a game changer. 

If you need to show some time savings elsewhere, I suggest taking a look at my company https://rippleengineeringsoftware.com/ which really helps you accelerate the heating and cooling load calculation process. 

[Iev60047]

Great! Thanks a lot for that helpful resource and advice, @Kevin.Lawson.PE ! Looking into the variables that I will need to define, I see that since my pump is in series operation, my flow factor will be 1 for both of the pump connectors.

Edit: After changing each of the connectors to match the kind of property that @Kevin.Lawson.PE provided in his document, I was able to set the set-up the piping system with two systems with only one system type (Hydronic Supply in my case), which is still one step from ideal (Hydronic Supply and Hydronic Return will have differing water temp's and thus differing pressure loss reports). So I need to change one Hydronic Supply system to Hydronic Return, but I can't figure that out. But I was able to generate a pressure loss report for both systems, that once added together approximated the pressure I tabulated in an excel sheet. Another problem I now have, is figuring out how to add the pressure drop for the return for the critical/index run for supply. And @Kevin.Lawson.PE , I may ask you some nitty-gritty questions about the details of the parameters set up for the Flow Parameter and Pressure Drop Parameter, but I am still figuring out what I need to ask. I think, I am getting closer to distilling a workflow for creating piping systems. One extra question I will need to answer is figuring out if detaching a component from a system and/or deleting a system is necessary, after I have altered the parameters of a component's family. I've been doing that, and it really ends up taking a lot of time to do (even for a system with just four components!).

[HVAC-Novice]

Not sure if this has been covered. but one thing I found not intuitive at first was that one closed loop consists of 2, a supply and return that each have a specific pressure drop you have to add to size the pump. 

One thing i do for device pressure drops is to plot the flow-pressure data in excel and create a curve fit equation. that I use to have the device calculate the pressure drop based on flow. So if the boiler has 100gpm, and for some reason I raise it to 120gpm, Revit will automatically correct for that instead of me manually entering the pressuredrop again.  I do the same for air systems. it is some work initially to do that for your diffusers, and all devices. but it will make design more error free and easier. 

[Kevin.Lawson.PE]

You only have one pump, so don't worry about flow factor.  Flow factor is for parallel paths in the system, so if you have two pumps in parallel, each pump gets a flow factor of 0.5, signifying that each pump gets 0.5*flow.  If you had 3 boilers in parallel you could set the flow factor to 0.33, ect. 

Yes, please, ask away, I'll do my best to answer although it has been some time since I've dug into this, I've mostly been focusing on load calculations and automated air terminal selection and layout. at https://rippleengineeringsoftware.com/

[iainsavage]

In both hydronic and air systems, for any component with fixed geometry, the pressure loss is proportional to the square of the velocity/flow so you can have a parameter for a reference flow and a reference pressure drop and then have a formula which will calculate pressure drop at all other flows, no need for excel tables or charts. Much simpler.

[HVAC-Novice]

It isn't exactly square and a device (i.e. boiler) has complex flow patters through varying openings, heat exchangers etc. For starters, a device with varying crossections will have varying flow velocities at any given time and flow will change from laminar to turbulent at different locations within the device. Some devices also include control valves or dampers that will move this far away from a square rleationship.

I use the manufacturer data and make sure "my equation" gives me the same or higher pressure drop for most if not all flows. 

Here just a random boiler I picked. the fitted curve gives me an equation that is not square at all and I do an error correction and manipulate the calculated values a bit to make sure they are not too low. For some devices the fit is worse. I assume the measured values are rounded somewhere and have some inaccuracy. the multiplier and exponent then will be used in the family to calculate the pressure drop based on the actual flow the device will have. 

Only the very simplest devices will be close to a square relationship. Even for a straight pipe, square relationship is just a simplification. 

[iainsavage]

Generally laminar flow would not occur in most hydronic systems unless velocities are very low.

Control valves or dampers obviously don’t have “fixed geometry” if they are moved to different positions, however in the fully open position, which is the case used for the index run and pump sizing etc, they will be fixed.

The whole theory of using K-factor or Kv value relies on the relationship of pressure drop varying with the square of the velocity and for any given fixed geometry flow passage velocity varies directly with volume flow, so pressure drop varies with the square of flow.

The boiler example that you have given is, in my experience, very unusual and I would actually be slightly wary of that data.

If you compare values for 11 degrees and 20 degrees dT in the example below you’ll see that it does follow the square relationship.

[HVAC-Novice]

For designing with a specific device, I use the manufacturer data and don't create my own theories. Manufacturers have instruments to measure that or to determine the design values with other means. 

In the rare case the published data don't make sense, I contact the manufacturer for clarification. For example, I came across odd data that turned out to be print errors.  

Many of the old methods are over-simplifications based on lack of testing, and computing power. And with enough safety built in, they can be used for design. But where available, I prefer actually measured data. I wouldn't take manufacturer data and say" they have no idea what they are talking about, I use my own data"

There are many things we don't consider. For example threaded pipe connections have higher pressure drop than welded pipe. If you have 2 fittings right next to each other they behave differently from being further apart. There won't be a perfect method. Add pressure-independent valves and other newer equipment and that whole square idea flies right out of the window. 

[RobDraw]

Thanks for sharing this conversation guys. I wish I had something of value to add but you're covering all that I know. So, I'm just gonna add my appreciation. 

Thanks!

[iainsavage]

Regarding the effect of screwed fittings, that should be built in to the K-factor and certainly in the “old” methods that I grew up with the K-factor for screwed fittings differed from that for welded fittings to take account of the pipe/fitting projection into the flow path.

Regarding PICVs, again you would assume that the device on the index run would be fully open and size the pump to suit, all other PICVs would then throttle as required to achieve the correct balance but unless you’re doing a full hydraulic analysis of the whole system you would not generally take account of that - the index run is the one that you would generally only be interested in.

Regarding manufacturer’s data they do not, as you assume, test all possible values. They usually test a couple of reference values and then extrapolate data for all other flowrates using equations, most notably P2=P1*((v2/v1)^2) and almost all of their charts will bear that out. I check that relationship in their charts before using that equation in my families. You’ll also find that manufacturer’s families also often include that same method.

Finally as an engineer rather than a scientist, I’m not interested in errors accounting for fractions of a Pascal in a system which has a loss of 50,000 Pascals or far more.

So I’ll continue to use my formula method in the 99.9% of cases where it is applicable and deal with the very few exceptions to that rule on an exceptional basis if and when they occur (after closely examining reasons for why they don’t appear to conform).
I think we’ve probably reached the point where we need to agree to disagree so I’ll bow out now.

[HVAC-Novice]

You are correct, fitting K-values account for threaded vs. welded. I stand corrected. Now I'm not sure if there are K-values for schedule 40 vs. 80 etc. So there is some inaccuracy. 

I'm talking about complex devices that have complex flow patterns, not simple pipes and fittings. If there are no data, the square relationship can be used (better than nothing).  But if the manufacturer publishes data, I use those. I also would contact the manufacturer if the information isn't clear or if it is suspicious. Some devices (i.e. heatpumps) have control valves built in - they are  a black box for us (at least for me). 

If someone can demonstrate my method of using manufacturer data is incorrect, I'm happy to change my ways. I always try to use data as correct as possible, but I try to err on the higher pressure side to be more conservative. 

You are correct, it all should be seen in the context of what % of the system pressure each calculation contributes. No need to worry about 1Pa. 

[Iev60047]

Wow! Thanks so much to @iainsavage and @HVAC-Novice for their valuable insights and knowledge of the up- and downsides to using Revit's built-in pressure drop capabilities vs. fitted curves vs. manufacturers' data for plumbing and HVAC systems. At risk of allowing other voices and opinions to pipe in and continue this general discussion of the pros and cons of different methods for calculating pressure drops in Revit (which I completely wouldn't mind!), I am going to leave my question unsolved for the time being. I am currently on vacation, but I will try to summarize what I have learned so far through trial-and-error and the input of others here into a response to include: (1) a general workflow, (2) some pointers, and (3) maybe the different decision trees to consider when setting up pressure drop calculations. As I have learned, there at least a few questions regarding this process that are not so clear cut as they initially appeared.

Edit: Numbering and formatting.