How to Model VRF in eQUEST

Script Preview

How to Model water cooled VRF in eQUEST

How to Model VRF Part 1

Hey, everyone. In this tutorial we’re going to cover the infamous VRF in eQUEST topic. There is no default VRF system in eQUEST, and thus we require a workaround. Knowing that, there are several workarounds, but of course you have to remember that they’re all workarounds, and even if you’re told what to do, that they are still workarounds and cannot be guaranteed to be accurate. We’re going to cover several methods here. The first method that we’re going to cover for air-cooled VRF is probably the method that you’re going to choose, because we made things quite easy for you.

The second method is a method that maybe you’ll choose at another time, because it can potentially yield and illustrate higher savings, but we leave that entirely up to you to decide. I’m just going to show you how we model VRF systems here, and you can make your decisions on how you want to model that.

First thing we’re going to do is create a new project form the wizard. Real simple. Open the DD wizard. I’m going to change this to something that’s got heating and cooling. Maybe something with a reasonably mild heating and cooling climate, like Denver. The only potential issue with Denver is that it’s so dry, but that shouldn’t impact our VRF modeling here. Leave it as an office building, that’s fine. We don’t need to worry about the cost — that will not impact the system type — and we’ll continue to the navigator. 

Here we just have a shell, and we have a system. By default the shell is going to be two stories, five zones apiece, so ten zones. That’s good. Let’s just set up the system. The key thing here that we’re going to want to keep is the system per zone. So one system per zone. 

Cooling is DX. Heating is DX. There’s several heat pump sources that we could have. We just want air. We’re going to be switching it to a packaged variable volume, variable temperature system. So we’ll just select the PVVT system, but that will change it to system per floor. We want it to be system per zone. We can set up our parameters. I’m just going to leave everything at the default because that really won’t impact how to model the system. We go to finish. 

So there are several ways, if you search around on how to model VRF — actually it’s pretty hard to find even a way to just search around through a search engine — but, one of the popular ones is published by Daikin. That’s the one that initially I started to follow, but the methodology is really, really slow. So if you go search Daikin VRF guide, you can find this modeling guide. It’s a PDF, here. Twenty-two pages. It comes with the download available — not with the PDF, but on their website, they have a BDL library data file. They want you to swap that out with your existing eQUEST data file, or BDL library. Really, all they’re doing is installing some customized VRF unloading curves into the default library, so that they’re available in eQUEST. Then what you need to do is follow through this process, and you can see it gets kind of long. This is just what you have to do to swap the files out, we’re not even to modeling anything yet. Not taking anything away from them; they did certainly a very thorough job of adding this, but the method here, for myself, made it easier. 

For every type of Daikin system, there are six unloading curves, and we have to select them and bring them through here. We’ll see what that looks like in eQUEST in just a moment. Every time you have to model a particular system, you have to go and select these curves and redo this, and redo that. I thought that was a waste of time, so I came up with my own way.

We just go to the component tree. Airside Systems. We left plenums in the building. We could’ve taken the plenums out. They’re going to be unconditioned spaces anyway. But you see we have a systems space. I didn’t even name anything, we just have ten systems and ten conditioned zones with each assigned space. If we wanted to install this, we could go into our system. Let’s go into detailed mode. We could edit our system from here. We can see, we have a PVVT system. Everything’s default in here. Heating is a heat pump. The systems type is PVVT, that’s the typical option for a VRF workaround. 

So instead of going through there, we could have the BDL library installed, and import all of the curves into each model. We’d have to do that each time. Alternatively, I created my own files from those curves. I’ll open the window here — tabbed over to Notepad++, which is the preferred text editor that I use. This just has some system input, but what’s in here are the VRF curves that came from the BDL library. I had to change the formats, but it’s all the same information. That’s just some common eQUEST knowledge that the BDL library has a different format than a .inp file. I wrote them up by VRV type for the Daikin systems. So this is the VRV III, they have a VRV IV — they have a couple options. VRV IV is with the heat recovery and a variable rate temperature, or with a heat pump, and there ends up being, I think, five options. 

Here are the curves, and rather than bringing these in one at a time through the BDL file, if we convert them into .inp format, we can import them all at once. Now, I went a step beyond that, and I set up a default system, and have the curves set up already. It saves quite a bit of time to do it this way.

What I did here, we have this all set up, and I just go to File, and I put the file into my eQUEST folder. File > Import File. We’ve discussed this in other videos. I have my VRV import files here, and right here I just import the system. We’re going to pre-process it, that’s just checking for errors. In order to make this move smoothly, I did have to put in a dummy space and a dummy zone. We could go ahead and delete that. Here I have the system. We’re in detailed mode — let’s just look at the spreadsheet mode. VRV III. Scroll through here. Everything is basically the same. I put the sizing ratio at 1. I could edit the input file and have it at 1.15, and it would input the import with 1.15. 

If we just look at the cooling unitary power, there are a few defaults that I did put in. Let’s look at our capacity curves, here. You see the capacity curve brought in the Daikin curve. Each system type has a different capacity curve by Daikin, so each one of the several system types does have a different capacity. If we look at the heating coil capacity, everything’s the same there. The heating capacity curves though, we’ll see that there is a VRV curve for that. That’s just two of the six curves so far. 

If we go to cooling, and unitary power, you can see for the VRV system, there’s an additional two types in cooling, and that would be the same for heating, so we have six curves. The curves are multi-variant, so there’s a lot of data points in the curves. If we just open up the system, there are only a few factors that are significant here. One of them is the fan. I brought in a default fan. The static — if the fan was ducted, the static would be something about .5 — typical watts per cfm, so for kW/cfm would be .0003.

For this, because we have one system per zone, when in reality you might have your zones larger and we could have several VRF units being estimated here, you’d want to determine the weighted average kWs per cfm and put that in here. These are the main inputs that you would get from your product spec sheet, and this would be different from project to project. 

The other main factor that’s going to change is the cooling unitary power input and the heating unitary power input. That’s the EER or COP. For this one we have a COP of approximately 4, and that’s just the inverse because it’s Btu/Btu in the EIR rate. That would, again, be a weighted average if you put that in. These values are available from Daikin, usually on the cut sheet, and you could take a weighted average. In that case, instead of by CFM, it would be by ton, by cooling ton. You’d take the weighted average, or by cooling Btu. So the larger unit would have more weight on the total efficiency. Alternatively, if there was a sequencing issue, you could figure out a new weighted COP from that as well. Keep in mind that units that you’re going to install on a project aren’t typically going to vary by that much, so you should probably be able to come up with a number for this that’s pretty accurate.

From here, you can see mostly everything else that I have is in green, and that means that it’s just defaults. The fans are pretty important, and if you note, for the fans by default, these are variable speed. If you watched our water cooled VRF tutorial, we mentioned that the fan control has to be constant volume in that case. In the air-cooled scenario, they can be variable speed fans, and of course that’s the major benefit here. Really, just a matter of importing the system, setting up the fan power for the supply fan, and whether or not there’s a static pressure, if it’s a ductless unit or a ducted unit, it’s of course an issue. Then the design of kW/cfm — these other numbers are an issue as well, but we left the defaults. So just the fans, and then you want the cooling unitary power and the heating unitary power. You can leave the curves, you could edit them, in theory. You really ought to, but I know that nobody will. Maybe one out of a thousand people.

And then of course, we do want to decide if the fan is variable speed or constant volume, and that’s really all there is to setting up the system. Now we have to set up the rest of the backbone of the file, and that’s where this could get a little tricky.

Now that we have all of these curves in here, what we could keep doing is, we could just edit the .inp file and make ten systems with all different names. So, VRV IIIA, IIIB… or, we could change the name of this system. We could edit this, change the name, re-import, and now if we go to VRV Import Files, we can pre-process it again, and since we changed the name, it didn’t pre-process over that name. We got an additional system type that is identical to the previous one we did.

However, my thoughts on that are: what you probably going to do is you’ll import this and then you’ll set that up according to the three tabs that we said you ought to change. So instead of that, you probably set up the system, create another HVAC system, copy an existing component, and then you’d want to give these good names. Maybe if it’s the Zone Type II or III or IV. And we just copy that — and if we had set up the parameters, we continue to do that. Create another HVAC system. Now you could see this still is cumbersome. Now imagine having to select six unloading curves each time we did this. Twelve, because you’d have to import it into your file first.

Let me just copy an existing component. We have four. If you’re good with the .inp file, you could figure this out ahead of time, and you could just copy and paste the system, change the name, and use that dummy zone over and over as the control zone. It would import the file that way. So, if you like copy and pasting a .inp file better, that would work. I don’t want to confuse anyone by trying to save a few minutes. 

What I don’t really like about this method is that these are not actually sharing heat between the systems. There’s no communication between these systems at all. They’re just simply relying on the unloading curves to assume standard efficiencies at certain loads that you naturally get with VRV.

There we have it — now we just have to go through this and painfully set these up. Alternatively, we could go into a spreadsheet. We also want to remember to delete the dummy zone, but we actually can’t do that quite yet. We have to remember to make the control zone. We have to remember that these were in order, and my nomenclature was a little skipped up here. What we can do here is, the first one is EL1 South Perimeter. The second is EL1 East Perimeter zone. Third one is EL1 North Perimeter zone. The fourth one is the Core Zone — five, five was the Core Zone, and four was the West Zone. That’s what I get for naming things poorly. My one mistake of hyphenating and then not, as I was hurrying — costing me more time here. 

Five was EL1 Core Zone. Six is South Perimeter zone. Seven is East Perimeter zone. Eight is North Perimeter zone. Nine is West Perimeter zone. Ten is Core Zone. 

From here, what I can actually do is delete the dummy zone now. Probably want to delete the dummy space first. Unfortunately, that was the only way that I could get it to import without being really annoying and then giving you an error, forcing you to assign something to it — so let’s just delete that space. Then Air Side HVAC, delete the zone. 

If we did everything correctly, we shouldn’t have a problem deleting the dummy zone. Otherwise the dummy zone was the control zone for every system. We should be able to delete these. I actually don’t think we have to delete them, but just to clean up the model, it’s easy enough. So there are other things we could have done. We could have copy and pasted the loading curves into these PVVT systems, just based on our spreadsheet from the imported systems. So there’s multiple ways that we could’ve done this. I’ll save one of these just to show you what else we could have done, rather than creating new systems. 

Here we have the one blank system left. It’s a PVVT. If we looked, everything was basically the same. I did change the sizing ratio, it’s really no big deal. That’s something that we could edit in the .inp file, so if you recall, we had the cooling coil capacity. Actually, everything was in the capacity curves were different until right here. We could just set the VRVAC. You have to make sure it’s the same one, this gets very confusing very quickly. So we could just set that, and I scroll over. Everything’s the same there. Then it would be the cooling unitary power — .2565. We can just use Ctrl+C and Ctrl+V. We could have done that for all of the other systems, then we would have to do the same for the heating coil capacity curves. 

Alternatively, we could have just changed the number of these rather than creating them or deleting them. Depending on how many systems you have in your file, that would be up to you. We still need to go to heating unitary power. Efficiency was definitely different. We can copy and paste one of these, but you’ll notice this is the heat PLR. For whatever reason, when I copied and pasted, it pasted the wrong one, and you definitely don’t want to do that. I tried to copy and paste both of them at the same time thinking it would work, and it did not. So, you definitely would want to look through and see that it was consistent because that would be a major discrepancy in the two curves. 

That would be another way to set up the system. So we could basically just copy and paste all of the inputs from the imported system, which maybe that would have been faster. It’s hard to say. Definitely would have been faster if there were 20 or 30 systems. I’m going to delete this one anyway. It probably would have been faster, because then we would not have had to change the control zones, but you live and learn. I typically would have done the opposite method and I would have probably just copied a bunch of systems from the .inp file, but alternatively, the copy/paste method would probably work pretty well. 

You’d have all of these systems set up as VRF. The main problem with the copy/paste method is, we are assuming these are all VRV III systems right now. If we had a mix of different Daikin units, VRV III, VRV IV, VRT systems — we would have a whole different set of unloading curves, and we’d have to identify each one of these one by one, and we’d also have to identify the efficiencies for cooling and heating one by one, and so on and so forth. So, it wouldn’t be that easy. 

We could easily make some mistakes any time we change the fan, or anything like that, from the default systems. The method that I showed is probably going to keep you a little bit more on your toes, in terms of accuracy. From here, we can simulate, and we have a workaround for VRF models. That’s all there is to that. The next method we’re going to show you is actually, in my opinion, a better method. I don’t really think many people will use it because it’s a lot more difficult, but it should be more accurate. In fact, it’s more accurate and it probably doesn’t save as much energy. Looking forward to that. 

Hope you enjoyed this tutorial. You can find these unloading curves when you sign up for the course. They should be available on the course page. Alternatively, they may be in your downloads folder at your user account. Okay! Good luck with the modeling of VRV systems.