• Home
• Cycling
• GP-Sana⇗
• Travel Blog
• Zuba
• tBuild⇗
• Lochaber⇗
• About

Mitsubishi Zuba Cold Temperature Heat Pump

Note: see new section on Fuel Comparison: 2022 vs 2023.

Contents

1. Objective
2. Background - the house
3. Zuba Furnace
   1. Controller - in hallway
   2. Zuba Temperature Sensors
   3. Supplemental Temperature Monitoring
4. How does it Work?
   1. Heating - No setbacks
   2. Setbacks and Disabling Heating Coils
   3. Energy Save Mode
   4. Electric Heating Lockout
   5. Humidifier
5. Data Monitoring
6. Specification Interpretation
7. Fuel Comparison: 2022 vs 2023
8. Reference Documents
9. Revision History

1.0 Objective

Our Zuba heat pump (PVA-A36AA7 + PUZHA36NKA) was installed in the summer of 2022 replacing a 95% efficient 18 year old natural gas furnace and a Trane air conditioner. We also replaced a natural gas hot water tank (low efficiency with chimney) with an electric hot water tank.

So far we really like our Zuba but we found that the documentation on how it works is sparse. This website is an attempt to describe how it works in our house supported with data collected on temperatures and power consumption in the fall 2022. The focus will be on the heat mode of operation. (Cooling works great too!)

Caution: This document is pretty detailed. Some people may find this pretty dry reading.

2.0 Background - the house

• Our house was built in the 1920s and is a mansard roof design.
• An addition in the same style was added to the house in the 1980s putting the area per floor at 600 sq ft. We have a total area 1200 sq ft when you ignore the basement.
• Windows are about 20 years old and many of windows are double thermal Argon E mixed with some double glass Pella with E coating.
• The doors are new (fiberglass and steel) and the pine sliding door glass has recently been upgraded to double thermal Argon E.
• Attic insulation was upgraded to R32 about 10 years ago. Basement walls were insulated to approximately R15.
• A Home Energy Efficiency Evaluation was done on the house in 2022 (for Greener Homes Energy Grant) putting the current house load at 12.7kW and recommending upgrades of: furnace, hot water tank, windows, attic (from R32 to R38) and air leakage sealing .. in that order.
• The natural gas furnace was a York Diamond 95% efficient 2 stage 55,000/36,000 BTU/hr which was adequate. Hot water was gas with chimney exhaust. Chimney was sealed after removal in summer 2022.
• We also have a Jotul gas stove (15,000 BTU/hr, 70% efficient) in our small living room and use it periodically. The room has lots of windows and tends to be the coldest room. We plan on keeping the Jotul as supplemental heat and emergency backup. We have a kit to convert to propane which is an option if we want to remove our gas line someday to remove the monthly service charge.

3.0 Zuba Furnace

We installed a Zuba Cold Temperature P-Series heat pump (PVA-A36AA7 + PUZHA36NKA) with auxiliary electric heating coils and air handler. It is a 3 ton and is spec.'d at 38,000 BTU/hr down to -15C outside temperature. 80% of that down to -25C. Each of the two heating coils deliver 7.5kW (25,500 BTU/hr ea.) when activated. (see spec page .)

Air Handler

Heating Coils H1 and H2

Outdoor Unit

3.1 Controller

Controller

The temperature control unit is small, simple and easy to read. It lets you set fan mode, heat/cool mode as well as target temperature (To) set point. You can easily see current temperature (Tc) as measured by sensor in the unit and set 'hold mode' - over riding any scheduled setback programs. All is good until you set up a setback schedule and that is when the user interface is poor - but adequate.

We put our controller in the hallway which is central to house and away from heat sources and vents.

3.2 Zuba Temperature Sensors

The Zuba room temperature measurement comes from the sensor in the controller (Tc - bottom right corner).

TRA - Return Air temperature sensor

Another temperature sensor is built into the air handler in the fan box and is used to decide when to enable the auxiliary heating coils. It is called the Return Air Sensor (TRA).

There are other sensors in the outside unit to decide when to go into a defrost cycle as well as a sensor in the heat exchange coils in the air handler. There may be others.

3.3 Supplemental Temperature Monitoring

Plenum Temperature Sensor
(exit air temp.)

Labjack data logger

To better understand how the unit works, a temperature data logger was added to measure outside temperature, temperature of plenum close to the outlet air, temperature in the hall close to the controller and temperature in the coldest room with the Jotul. All measurements recorded and averaged to once every two minutes then combined with hourly power consumption data from our utility company.

4.0 How does it Work?

The following images describe how the Zuba controls the H1 and H2 heating coils based on different TRA (return air temperature) and To (set point) scenarios. These images are from page 21 and 22 of the P Series Service Manual (item 1.3 on spec page). Click on images to open for zoom.

Logic table H1,H2 activation sequence - p21 (click on image to open)

Graph showing H1, H2 activation sequence - p22 (X = 24 minutes) (click image to open)

4.1 Heating - No setbacks

The simplest way to operate this unit is to set the desired temperature of your house with no temperature setbacks and let Zuba do the rest. This is what will happen:

• Zuba will maintain the temperature within 0.5C of the set point (To) using the temperature sensor (Tc) in the controller.
• Zuba will use the intake air temperature (TRA) to decide how much heating power you need and adjust the heat pump power accordingly. It does this as follows:
  • If the temperature difference is 1.5C or more between the set point (To) and the intake air (TRA), it will turn up the power on the heat pump. (this is a guess - can't find documentation on this.) Measurements suggest there are at least 3 levels of heat pump power.
  • If after 24 minutes the difference is still 1.5C AND the intake air temperature has not increase by at least 0.5C, the 1st electric heat coil will be turned on adding 7.5kW more heating power.
  • If after another 7 minutes nothing has changed (still 1.5C diff and no 0.5C increase), the 2nd heat coil is turned on adding another 7.5kW of heating power for a total boost of 15kW or 51,000 BTU/hr on top of what the heat pump is delivering.
• In all cases, if you leave the fan in auto-mode, the fan speed will increase as more heat is generated in the air handler. The algorithm for this is not clear but it's probably a heat sensor in the coils as a minimum. Perhaps the fan goes on high when the heating coils are turned on too.
• If the temperature difference between TRA and the set point (To) is LESS than 1.5C, I believe the unit uses the controller temperature sensor Tc (in the hall) to decide when to stop delivering heat. It is unclear what the algorithm is to increase or decrease the power of the heat pump when this (To-TRA) is less than 1.5C and whether it continues to use TRA (inlet air temp) in some way. Clearly, the unit must eventually use the hall controller Tc measurement to decide when the house has reached set point.

4.2 Setbacks and Disabling Heating Coils

If you never adjust your set point temperature by more than 0.5C, you probably will never see the electric heating coils come on unless the heat pump can't produce enough heat when the outside temperature is very low (below -15C?).

If you program a setback to reduce the temperature in the house at night, you will likely see the electric heating coils activate when the setback is removed in the morning.

In our case, we decided to set the temperature at 21C during the day and at 9PM schedule a setback to 19C for sleeping. We decided to start ramping the temperature up at 5AM to 20C and then 6AM to 21C in the hopes that the house would get to 21C by 7AM when electricity rates go from the lowest rate to the highest rate. We did not want the heating coils coming on so we flicked the breakers off to save more energy and tolerate a longer return delay to a house temperature of 21C.

Note: If you don't want the heating coils to come on, you can turn off the heating coil breakers (H1 and H2 .. see photo below). If you are not worried about freezing pipes or are always present in your home, you can turn on H1 or both H1 and H2 when you realize the heat pump is not keeping up with the set point ... or if you want your system to respond faster to changes in set point. You can also install a low temperature electric coil controller - see Electric Heater Lockout later in this document.

H1 and H2 breakers

Note: H1 is larger capacity because in some installations the fan is on this circuit.

4.3 Energy Save Mode

At first I thought that Energy Save was how to manage the activation of H1 and H2 heating elements .. but I soon learned this is not so.

Energy Save Mode, from what I can understand, is a way of limiting the heat pump power consumption during periods of the day when the cost of energy is high. It has no knowledge of H1 and H2. I was also told it only works in Cool mode not Heat mode. I'm not sure that's true .. but it certainly does not know anything about H1 and H2.

The Controller Instruction Book on page 24 is probably the best place to look .. bottom of page. ( See item 6.2 : spec page) I think there is another place that talks about the percentage being of the rated capacity of the unit. Yes you can set 0% and shut the unit down.

Someday when I have too much time on my hands I'll make measurements on this.

4.4 Electric Heating Lockout

If you decide you want to use setback programming to lower the temperature of the house at night, you may want to disable the heating coils from coming on in the morning .. provided you can accommodate the slow return to your day time target temperature. (It can easily take an hour to raise the house temperature by 1 degree C.)

The caveat to this is if the outside temperature is low enough to make this recovery time excessively long or perhaps even impossible. In this case you will want to enable at least one if not two of the heating coils.

The solution is of course to live in your house and know when it will need the heating coils based on outside weather and simply flick the breakers on.

The other way is to install a controller that measures outside temperature and decides when you will need to enable H1 and H2 to keep the house warm. Mitsubishi has an add-on you can install to do this and it is described in the Electric Heating Lockout section (5.0) of spec page.

The Mitsubishi Lockout solution uses a Ranco 2 stage controller with an outside sensor. You can set the temperature that enables H1 and a further (lower) temperature that enables H2.

Ranco - 2 stage controller

Penn - by Johnson Controllers

I'm sure this Ranco controller is good but we decided on the Penn Controller made by Johnson Controls. They only offer a single stage controller but we decided that we could enable both coils and let Zuba make the decision on when to enable H2 (7 minutes after H1 if no change in in-take temperature TRA). We added a relay to enable both H1 and H2 and selected the 24VAC version so we could share the humidifier transformer and keep the wiring all low voltage. You will see these items in section 5 of the reference documents.

Of course with this solution we can also keep the H2 breaker off to restrict the heating boost to 7.5kW (26,000 BTU) if we want.

We programmed it to enable the coils at -15C and disable coils when temperature rises to -12C. We are waiting for some cold days to monitor temperatures to watch this happen!

Note: I don't provide a circuit diagram because if you need one then you should probably not be doing this upgrade yourself. Ask your dealer to install the Mitsubishi solution.

4.5 Humidifier

Finally, there is the humidifier to consider. In our case we re-installed the flow-through unit from our previous furnace but held back on the wiring to see if we really needed humidity in the winter. We decided we DID need some humidity and so we wired the unit up with just the humidistat controlling when to turn on the water solenoid with no Zuba control board connection. We decided this was OK because the fan is running at low speed most of the time and the heat pump comes on frequently as the outside temperature drops - so there should be enough air flow to humidify the house.

Mitsubishi has an add-on (CN24RELAY-KIT-CM3) to work in conjunction with the heating system (to set fan on high when the humidistat turns on). It is described on page 25 of the PVA Installation Manual. ( See item 1.1 : spec page)

So far, our humidifier is working without the relay kit and is meeting our target of 35% relative humidity.

5.0 Data Monitoring

As stated earlier, we installed some temperature sensors around the house and collected temperature readings to better understand how the unit works.

Graph #1: 2022-11-18 - heating coils off - 9PM setback (click on graph to open for better look)

In the first graph (Graph #1 above) both H1 and H2 heating coils are off and the temperature sets back to 19C at 9PM and returns to 20C at 5AM and 21C at 6AM.

This is what you can observe:

• the heat pump did not come on after the 9PM setback until 1AM and then again at about 3:15AM. Notice the red trace (plenum temp) spiking down and back up again in the first heat pump session at 1AM .. that's the unit going into defrost cycle.
• At 5AM the set point is 20C and the heat pump power level goes up pushing the plenum temperature into the 41C range.
• At 6:45AM another defrost cycle (spike down then up)
• At around 7:50AM the heat pump power level drops (4kW-hr to around 3kW-hr) with a plenum temperature drop to around 36C from 41C. This corresponds to the hall temperature reaching the set point of 21C.
• Note also that the living room temperature is always at least 1 degree C lower than the hall temperature and it takes a bump up just before noon when the Jotul is turned on for a boost.
• Last points of interest are: the number of defrost cycles (5) and the 9PM temperature drop as the setback to 19C takes hold.

Graph #2: 2022-11-18 (click on graph to open for better look)

This next graph (Graph #2 above) was in October 2022 with H1 and H2 enabled. Outside temperature was much warmer as you can see and in this case the set point was returned to 20C at 6AM and 21C at 7AM (from a 19C setback at 9PM). (The living room temperature sensor has a problem so ignore).

This is what you can observe:

• At 6AM the set point temperature compared to TRA (air intake) is not sufficient to start H1 and H2 sequence. The heat pump pushes the plenum up to around 35C.
• At 7AM the set point is now 21C and the H1 element is almost immediately turned on. The plenum spikes up quickly to 49C and the hall temperature soon makes it up to 20C which turns off H1.
• Power numbers show the unit has used 5.3kW in the hour .. and H1 was on for about 1/3 of an hour. (This would add 2.5kW-hr to the heat pump of 3kW-hr giving us the 5.3kW approx. .. so makes sense and H2 never came on).
• We can't see the 24 minute delay after the set point changed to 21C .. is this because I can't remember exactly what the controller was scheduled to? .. or does the unit not know the history of the set point?
• Clearly, I need to collect more data on H1 and H2 engaging on a colder day .. so stay tuned!

6.0 Specification Interpretation

It took quite a bit of digging to find the P-series Specifications ( See items 3.1 and 3.2 : spec page) document and home in on the model combination we purchased (page 47). The M-Series Spec. was easier to find and talks about COP factor .. but the P-Series is a little newer and the M-Series is obsolete.

COP is the Co-efficient of Performance and is the units heating capacity (power output) divided by the power input. Of course you need to convert BTU/hr to watts to keep the units consistent. (3.41 BTU/hr is equal to 1 Watt)

The Learning Metrics Website ⇗ website has an interesting graph on a typical heat pump as seen below:

A Typical Heat Pump Coefficient of Performance graph

So let's see if we can generate a similar COP graph for our heat pump. With some cut-and-paste of page 47, we have a spreadsheet with COP calculation based on the Max Q/W (not Rated column - see note 1).

• page47-spec.xlsx ⇗
  
• page47-spec.pdf ⇗

It is interesting to graph these numbers to see how they compare with the 'typical' performance curves from Learn Metrics. You see (below) that there are 3 families of curves for Q, COP and W for each of the three room temperatures of 25C, 20C and 15C. The right Y-axis is for the COP curves and the center Y-axis doubles for Q (BTU/hr) and W (Watts). Of course the bottom X-axis is outside temperature. (Click on graph to open up in larger window!)

Heating Mode Spec. PVA-A36AA7/PUZHA36NKA (click on graph to open for better look)

So let's see if we can interpret these curves. The Q (BTU/hr) curves seem somewhat fabricated giving a flat section from +5C to -15C (outside temperature. The 20C room temperature line is on the 38,000 BTU/hr spec of the heat pump and it 'degrades' nicely to 30,000 BTU/hr on a straight line to -25C which is conveniently 80% of 38,000 BTU/hr. This fits the published spec of the Zuba unit perfectly.

Take at look at the W (Watts) input curve and you can see the power consumption of the unit is 3000 W at +5C outside temperature which gives you a COP of well over 4 which is great. The unit then requires more and more power to follow the Q line until it reaches 5600W at -15C (COP=1.83) and then 6000W at -25C (COP=1.42).

With a COP of 1.45 at -25C, it is still better (in theory) to run the heat pump over the heating coils (COP=1.0) indicating it is providing 1.45W output for every 1 W of input.

What is somewhat surprising is that the COP curve looks so twisted when compared to the Learn Metrics example. I suppose we have to remember that the Zuba is not a traditional heat pump and has no compressor. My guess is that the firmware running the unit is making a lot of decisions to squeeze out as much efficiency as it can resulting in a not-so-pretty curve.

With the data we are collecting on inside and outside temperatures, we should be able to see how close we come to the theoretical power consumption based on the COP data spec over this winter ... and compare it to our utility bill.

Fuel Comparison - 2022 vs 2023

The bottom line is, how does the 2022 season compare to 2023 after installing the Zuba heat pump?

We got quite a few people saying we were going to regret converting and that our energy costs were going to go sky-high. Fortunately they were wrong and in fact (spoiler alert) we came out almost 'at par' comparing natural gas with the cold temperature heat pump from Mitsubishi. Note however that 2023 was a much warmer winter compared to 2022 and the Zuba looses it's efficiency below -15C.

Here is the data as collected from Ottawa Hydro and Enbridge bills with analysis. A table of definitions follows.

Note: In Ontario, and probably many provinces in Canada, the term 'Hydro' means 'Electric Power' which historically has been hydroelectric power. In Ontario, a large percentage of power generation is nuclear power (54%) , hydroelectric at 26% and natural gas and wind generation at about 10% each (as of spring 2023).(See ieso.ca ⇗)

Notes:

1. includes all energy consumption (hot water, stove, lights etc)
2. March and April the heating is minimal (travel months)
3. we still have gas to a Jotul heater and it was used periodically on cold days
4. In 2023 we had 19 days with a temperature minimum below -15C (point at which we are enabling one heating coil) (That # was 47 days in 2022!)
5. Hot water was on gas in 2022 then electric tank 2023
6. clearly the energy audit is not accurate putting us at double the consumption (compare line #11 with #19)
7. 2023: 10% warmer winter, 30% cooler summer (lines # 15 to 18) but especially look at # days below -15C which will have a more significant impact on heat pump efficiency (to the down side).
8. dollar figures are typically within 5% of actual due to fluctuating rates and charges by Enbridge and Ottawa Hydro

Some interesting Stats from the web:

• Natural gas pilot light typically consumes 600 BTU/hour (we turn ours off when not in heating season)
• Typically 15% of energy consumption goes to hot water heater
• 4.2 metric tons CO2 for 2 people to fly Montreal to Valencia Spain (return)
• Toyota RAV4 AWD generates 600kg CO2 per 1000km – or 12 metric tons per 20,000 km (ie. per year)

Some Conclusions?

1. Our Zuba moved us from 1.20 to 0.18 metric tons of CO2 (line# 12) and did not appreciably cost us more to operate. We would save even more if we did not have our Jotul natural gas stove!
2. We used half the energy (GJ) with Zuba compared to natural gas the year before .. but since electricity is twice the cost per GJ we came out on par.
3. Green Energy Audit did not provide any meaningful numbers (Even Heat Load is suspect).
4. Monthly service fees for both natural gas and hydro can add up quickly and in our case cost us an additional $300 per year. However, that's our backup plan if we have an ice storm and loose power. We also like the warmth of a gas stove on cold days.
5. Our Toyota RAV4 and trip to Spain are our major CO2 contributors and heating our house is not and never was a big CO2 contributor.
6. We need another winter like 2021 to do a better energy comparison. Such a winter will push hydro costs up for sure.

Definitions and Conversions

• 1 MMBtu = 1 million BTUs (energy)
• 1 MMBtu = 1.054 GJ (giga Joules) (energy)
• 1000 cubic metres of natural gas = 36.9 MMBtu (volume to energy)
• 1 Watt = 1 Joule/second (power)
• 1 Watt = 3.412 BTU/hour (power)
• 1 KWatt-hour =3412 BTU = 0.003412 MMBtu (energy)
• 1 Therm = 0.1 MMBtu of NG
• 14.43 kg of CO2 from 1 MMBtu of natural gas (CO2 emission)

Reference Documents

Most of these Mitsubishi documents were downloaded from https://mylinkdrive.com/ ⇗.

PVA-A36AA7 (air handler, coil assembly)

• 1.1 Installation Manual: M_PVA-A-AA7_Install_MD-1404-K024_January_2022.pdf ⇗
• 1.2 Operation Manual: M_PVA-A~AA7_OPERATION_MD-1404-K006_2018-04.pdf ⇗
• 1.3 Service Manual: M_PVA_MD-1404-K011_V-3_Service_2-22.pdf ⇗

PUZHA36NKA (outside unit)

• 2.1 Installation Manual: M_PUZ-HA-NHA1-NKA1_INSTALL_BH79D886H03_10-20.pdf ⇗

Specifications

• 3.1 Technical Data Book: M_PUZ-HA-NHA1_NKA1_TECHNICAL_DATA_BOOK_OCD750B.pdf ⇗
• 3.2 PVA-A36AA7 / PUZA36NKA spec (page 47): M_PUZ-HA-NHA1_NKA1-DataBook-page47.pdf ⇗

Electric Heat (15kw)

• 4.1 Installation: Electric_Heat_Kit_Install_Instructions_2-22.pdf ⇗

Electric Heating Lockout

• 5.1 Lockout Controller: Electric_Heat_Lockout_Controller_PLG_23260.pdf ⇗
• 5.2 Lockout Install: Electric_Heat_Lockout_Install_Instructions_03-09-15.pdf ⇗
• 5.3 Ranco 2-stage controller image: ranco-2.png
• Johnson Controls:
  • 5.3.1 image: penn-A421GBF-02C.png
  • 5.3.2 Installation: penn-A421_24-7664-3019-InstallGuide.pdf ⇗
• 5.4 Dig-Key 24vac relay: relay-24vac-dpdt.png

Control Unit (PAR-CT01MAU-SB)

• 6.1 Installation Manual: WT08632X01_IM_PAR-CT01MAU-SB_MEUSA.pdf ⇗
• 6.2 Instruction Book: M_PAR-CT01MAU-SB_OPERATION_WT08631X01_2017-11.pdf ⇗

Notes

1. From a comment on https://hvac-talk.com website: "Ignore rated capacity, it's basically an engineer geek thing that doesn't mean much practically. That number is taken when the compressor is locked at a fixed speed to put it on a level playing field with non-inverter units - completely wiping out the purpose and advantages of the inverter.
   Maximum capacity is the number you want to look at."

Revision History

• 2023-01-02 - first release
• 2023-07-09 - added section on Fuel Comparison

photos by Harold Blount - Copyright © 2007-2024
hb at nordicwind dot ca
legal