How do you model minimum and high efficiency DX cooling and heat pump equipment?

savecancel

Generally, for smaller DX units, I've never had any luck getting detailed info from manufacturers, and sometimes it's even hard to get design COPs for really small units. Generally, when I have to make assumptions, I stick with 90.1 baseline efficiencies and curves (or just software default curves). That's not quite right of course, but it's a tiny impact.

Heat pumps tend to be more important, so I'll spend more time hunting down performance curves. It's also not as straightforward getting detailed performance curve info especially in the format needed for software (that has been adapted from the polynomial curves method for chillers), but I find it's usually pretty doable to get at least one curve for full load COP by OA temp (or source temp for WWHPs), and also another curve for capacity at those conditions - but not always possible/easy to get the full curve. because of that I'll confirm that there is plenty of capacity in the coldest conditions, and potentially ignore the capacity degradation in the model, while being sure to account for the temp dependency.

There is a wide range of DX equipment. My experience:

CRAC units with DX cooling coil for data centres - The manufacture Vertive provided all the necessary data including a performance table with various cooling conditions. The data are project-specific. Performance curves were generated by regression analysis using Excel.

VRFs - Techical data books with performance tables are found online for Daikin, Mitsubishi, Toshiba, Panasonic, etc. Performance tables for outdoor units are shown in all the technical data books. Performance curves can be easily generated by regression analysis. Performance tables for coils in indoor units are often not shown in the data books. When capacity modifier curves for indoor units are unknown, I use the same curves as those for the connected outdoor unit i.e. the change in capacity of the indoor unit is considered to be directly proportional to the change in capacity of the outdoor unit. This is the case with EnergyPlus. IESVE can't customise VRF performance curves. The necessary information of indoor fans is more accessible as it is in the product catalogues.

I seldom use the default curves in simulation software as they perform much worse than the actual product's curves.

I think about this in 4 main steps:

1. Determine rated efficiency & capacity
2. Determine fan power info
3. Determine efficiency/capacity after subtracting fan power
4. Review/adjust part-load performance

Step 1 is usually straightforward.

Step 2 is usually easy enough to get reasonable data for fan power at various airflow and static pressure conditions. Smaller DX/HP equipment doesn't always provide this. However, I can't think of any manufacturer that provides fan power data at AHRI rated conditions (and AHRI doesn't require separate measurement/reporting), so it's not technically correct.

Step 3 applies some formulae that appear to be commonly used, but old, varied and can't possibly be accurate across a wide range of equipment. I posted a question about this last year here. It's pretty disappointing that we still don't have a reliable industry-wide solution for this.

Step 4 is something we rarely do because it's time consuming and hard to get good enough manufacturer data to do it right. The idea would be to compare the efficiency/capacity calculated by one or more default performance curve sets to manufacturer provided conditions are various part-load points, and adjust curve coefficients so that they align more closely. The curves are provided in the software we use, and I believe are from the same collection that's been floating around the industry since DOE2.2 was developed. This is an increasingly important issue because codes such as Title 24 are getting more stringent performance-wise that designs need to get creative on the types of DX and HPs selected. If you select equipment with an invert-driven compressor and EC motor, you'd rely on the performance curves to reflect those benefits across the part-load range. However, the curves are pre-selected based on equipment type from a dropdown menu. There's no way to reflect those savings when there's one curve set and you can only adjust the rated efficiency. Even part-load efficiency like IEER is a user input to the software, yet is only there to check for compliance and doesn't actually affect the simulation.

EDIT: Side note that VRF introduces even more issues than I described above, and it's laughable how much the "generic" EnergyPlus curves that most software tools use vary from manufacturer published data or manufacturer-provided curve sets.

I asked this question over a decade ago to a couple of different suppliers in the UK. What intrigued me was the very high efficiency that suppliers (e.g., Mitsubishi, Daikin) quoted - figures that we just plugged into building code calculations. Based on these figures it seemed like a no-brainer to just rip out centrifugal chillers and replace them with acres of DX systems! My enquiries came to little. The only interesting thing that I got (from a Danish academic) was that the market was very Germany-centric which resulted in SEER & COP being established at what seemed like, for the UK at least, quite extreme temperature bins.

The impression I got is that, locally, there is some design experience within the sales teams but it's limited to sizing. The best I got back in terms of energy was something along the lines of "we've got some boffins over in Japan, we can try asking them."

I don't doubt that they have some very clever people. Are they clever enough to get centrifugal-like performance out of a scroll compressor? I don't know. Reading your message, it's interesting to see that maybe the US market is subject to the same "black box" issues!

Similar experience like Ben. For smaller DX units it's really hard to get the "curve" or "performance table". But generally, it's not hard to get the rated condition efficiencies and its power draws on compressor, condenser fan, and evaporator fan. I always model those components separately and adjust compressor only COP based on rated data, not the ASHRAE 90.1 equation. Then I would be using the software default curve based on its sizing category. The tool would then adjust it from rated to design conditions.

For HPs, it's a lot easier to get a "performance table" which will show the relationship between OA(or EWT for WSHPs) vs TC vs Power Draw. I then manually create curves without the fan because I will be modeling them separately. Note the curve would reflect compressor only because fans are out.

Note I pretty much never use the efficiency adjustment equation for my proposed design unless I really can't find power draws. I use it for the baseline, if there is one.