GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Toshiba VRF efficiency at part load

user-7651898 | Posted in General Questions on
Having tremendous difficulties with the system. It’s a 12 kW multi split VRF with four indoor units.
When running at full tilt it produces a cop of about 3.5. This agrees with consumer testing and more or less with the specified efficiency. Unfortunately when running a single indoor unit, especially for heating, the specified power input should be 700 W but it’s actually drawing 1500 W. This pull the efficiency down to 1.7 from claimed 4+.
My complaints to Toshiba and the supplier seem to be getting nowhere. They are insisting that “field results may differ from specification”. I agree this is the case in principle but being so far off renders the units useless for its intended purpose  of displacing oil heating.
I’m in the southern UK, would anyone know if there is an engineer I can contact to get a report commissioned? Any thoughts on the above?
I’m a physicist and very familiar with test instrumentation. I used a professional clamp meter  to measure power consumption. I’ve taken the power output as read and haven’t done a plenum test yet.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Expert Member
    Akos | | #1

    This is a very common problem with all multi split systems. The issue is that most (all?) bypass some refrigerant through the zones that are off. Most likely the one wall mount is bellow the min modulation of the compressor which kills COP.

    Minisplit manufacturers don't post this information anywhere which catches a lot of people out. Combine this with oversizing and associated cycling issues and you'll rarely see nameplate COP.

    Generally the only solution is to combine some of the small zones onto a single larger ducted unit that is above the min of the outdoor unit or replace the multi split with a number of one-to-one units.

  2. user-7651898 | | #2

    Thanks. I suspected that. The indoor is 2.5kw vs 12kw outdoor and the modulation floor is 30%. However it also does it with a 3.5kw indoor, showing 2.3 COP vs 4.5 spec and 3.9 Eurovent tested.
    My suggestion was to add an additional 3.5kw unit to bring the minimum load to 50%.
    That said, it just lost its gas for the 2nd time in 12 months … I’m inclined to reject it completely.

  3. Expert Member
    Akos | | #3


    Lets step back a bit. A 12kw heat pump, even in my pretty cold climate is huge. It will take a very large home that much heat especially in the southern UK.

    I would run through the calculation here first:

    You can use last winter's bill and assume a COP (2.5 or so is good guess) or much better, use a winter bill before the heat pump was installed.

    Once you have that, you can figure out the outdoor unit you should start from. You can then look at the best indoor configuration to connect to it.

    Generally with a multi splits what works the best is a single larger wall/floor/ceiling mount in the living space and a ducted unit to feed the rest of the smaller rooms. I've speced a couple of setups like that and they work very well in terms of comfort and efficiency.

    If I had to pay UK energy rates, I would skip the multi splits and go with one-to-one units. They generally have much better modulation range and higher COP. The BOM cost for a two zone multi and two one-to-one units is about the same, just a slightly increased install cost.

    As for leaking there is a small chance of pinholed indoor coil, the problem most likely is a bad install.

    1. user-7651898 | | #4

      Thanks Akos,

      In the UK we have an estimated energy consumption figure in the form of a pre-sale certificate produced by a surveyor. I used a simplified SAP calculation combined with degree-days and it comes within 10% of the actual oil consumption we use so I'm happy there. I also measure oil consumption with a logged system (Shelly PM if anyone is interested).
      I use a 36KW oil boiler, leftover from when the pool was heated from oil at 90% efficiency.
      Ironically, heating oil tracking crude prices is proving better than natural gas which will hit 12p/kWh in October and electricity at 40p.
      I have 12kW of PV south facing making 10MWh annually which pretty much covers the pool ASHP during the short UK swim season.
      So, I have an annual demand for space heating of about 25000kWh for a 3600sqft single story brick/block house with double glazing. That comes to about 7kW for a 5c winter average ambient and peak 14kw at -5c and windy.
      Heat load is higher as we're south facing and sometimes hit 30-35c these days. AC usually is on for 4 weeks a year.
      I have 2.5kw office (3kw load, large south glazing etc), 3.5kw kitchen (4kw load), 3.5kw master bedroom (5kw load), 2.5kw second bedroom (4kw load). Nothing in living area (5kw load) or 2 further bedrooms. These are cooling loads at 30c and full sun.
      So all in all I think 12kW is oversized for heating but not hugely given a cold start requirement and the lower output as the ambient drops but correct me if I'm wrong. It's the inability to modulate down that kills it.
      Hitting the office and kitchen gives a reasonable COP. Just the office though, terrible and this was the project requirement. The HVAC guy advised I take the big VRF over 2 separate 2-1 multi-splits.
      To say I'm not impressed is an understatement!
      I very much appreciate your help.

      1. Expert Member
        Akos | | #5


        It sounds like you did your homework. If the place does indeed have a 14kW heat load at design conditions than you have some serious low hanging fruit in terms of energy efficiency. Given your high local utility costs, fixing some of those might be well worth the effort. Won't fix the modulation issue but it would be cheaper to run.

        Going for a pair of two zone multi splits would definitely made a difference. Toshiba is not popular here so I'm not very familiar with their offerings but if it is anything like the Mitsubishi, the minimum on a two zone is about half that of a four zone unit.

        1. user-7651898 | | #6

          Low hanging fruit ... I wish it were so. I have:
          - cavity insulation
          - 300mm loft and all penetrations sealed
          - all renewed external doors, all composite insulated to latest standards
          - hydronic panel heaters but no mains gas so oil, all individually zoned with zigbee thermostats
          - low-E double glazing from 2012

          It's the house itself, very exposed location with 2 miles of flat fields to the prevailing wind direction, single storey and high perimeter/area ratio, huge area of glazing (which is good for the soul), solid slab foundation. Add in a South African wife and a teenage daughter and I'm on Greta's *hit list!
          When I bought in 2017 I considered demo and rebuild but calculated it would take until 2090 to pay back given generous factors.

          All I can think of with a reasonable payback, short of external wall insulation, is air tightness test and maybe MHRV.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |