I did quite a bit of testing over the summer of the Wave 2 that I purchased and thought people would be interested in the results. The short story is that out of the box, the bare bones $170 Toshiba is significantly more efficient in cooling capacity/power usage. Apologies for the long post, pulling in a lot of information from multiple posts made in other locations.
Background
The reason Ecoflow performs worse than its nameplate rating is because they use the ASHRAE BTU's which does not account for the heat generated by the unit being in the room (as opposed to outside with a window a/c), leakage through the ducts, and heat being radiated to the room from the ducts. The DOE recognized that portable a/c's have these limitations and started requiring manufacturers to run a new test with portable units to give an "equivalent" rating to the old test. If you look for portable a/c's you'll see two ratings now, ASHRAE and DOE SACC.
Unfortunately there's not a single way to compare the two numbers since there are so many things that effect the rating (insulation in the housing of the a/c, insulation of the ducting, single vs. dual ducts, air leakages, single vs. variable speed compressor, etc.). This page talks about the two systems and has some ranges of comparison: SACC BTU DOE vs ASHRAE in Portable AC - PICKHVAC
Using their ranges, it seems a proper de-rating can be anywhere from 50-85% to go from ASHRAE to DOE. Dual hose is the biggest differentiator, 75-83% de-rating per the website whereas single hose is 58-70%. We know the ecoflow isn't super well insulated and does have duct leakage problems, so a 75% de-rating seems appropriate, possibly even 70%.
That all boils down to the Wave 2 being equivalent to an ~3500-4000btu window unit but using the same amount of energy as a 5100 btu window unit. If you're not using both hoses it could be as bad as a 2500btu window unit.
What can we learn to improve cooling performance?
Test Setup & Unmodified Results
Over five days I compared the wave 2 and its standard modes (no modifications) against a 5000 btu Toshiba window unit that I had out in my garage. All tests were started at ~4pm and ran for one hour to try and get similar sun loadings across days. The garage door and back door were opened for ~30 minutes prior to each test to allow temperatures, thermal mass, and humidity in the garage to equalize to the exterior. The Toshiba has a slight advantage over the wave 2 for its test because I had previously built an insert with 2” of xps to match the window unit’s size. The Wave 2 only had 1” of xps and 1/8” sheets to finish closing the gap under the door. There was also some light rain ~20 minutes into the Toshiba’s test and my exterior thermometer got wet so there is some question in the final exterior temperature for that test only. The interior thermometer was unaffected. The garage is uninsulated (except for a couple panels on the door), and is ~2000 cu ft so definitely bigger than either a/c is intended for.
The Toshiba is rated for 4 amps at 115v AC (460w). I didn’t think to take power data during its test, but a separate 15 minute followup test using my delta mini showed that AC power draw was slowly ramping up to at least 400w. A longer test may have grown closer to the 460w rating. A short 5 minute test in low mode only used ~380w AC, but I don’t have any hour long cooling tests to look at temperatures in low mode.
In MAX mode, the wave 2 peaks at 470w DC. The wave 2 does do some aggressive throttling of the compressor during its operation in all modes. I believe it targets a delta temperature of ~18F between ambient and outlet temperatures in MAX mode. If the delta temp is >18F, it will ramp down the compressor, if it’s <18F it will ramp up the compressor. I didn’t record power data during the MAX mode test, but on a shorter 15 minute test MAX mode ranged from 440-470w DC. If you start getting closer to your set point I bet it would reduce further.
In ECO mode, the wave 2 cycles through a number of “modes” to save power. It starts with the compressor on high and low fan speed for 5 minutes. It ramps up over those 5 minutes to ~380w dc power draw. Then it switches to a low compressor and low fan speed for 15 minutes. In this mode it draws ~200w. Now comes the bad part, it runs on FAN ONLY for 10 minutes, only using ~10w. The cycle then repeats. I suspect that the same compressor limiting based on delta T could reduce the 5 minute high compressor numbers if you are near your set point. Overall, without compressor limiting, the ecoflow averages ~150w DC.
The issue with ECO mode is that the residual water in the base of the unit is re-added to the air during the fan only portions of the cycle. This is BAD for overall comfort. The RH in my garage actually INCREASED during the test which should definitely not be happening while an air conditioner is running. Perhaps a longer test would get to a better steady state and start to reduce humidity.
In sleep mode the aggressive throttling of the compressor actually made a difference in power consumption. The wave 2 targets a 20F (11C) delta t in this mode, and with the fan on such a low speed there’s significant mixing of the outlet/inlet air streams (and not enough mixing with the room air) so it ends up throttling the compressor down quite a bit. This has a significant effect on overall cooling ability of the room. At the start of the test the wave 2 was using ~250w DC, but within 20 minutes it dropped down to ~200w for nearly the remainder. Total energy use ~220w DC for the hour. I did run this test for an additional hour and room temperatures didn’t drop much further, but humidity levels did continue to drop (~1.5% RH/hr).
The last test, unmodified manual mode with high fan. The compressor throttling still occurred, targeting ~17.5F (~10C) delta t. I saw a max of 400w DC near the beginning of the test, dropping to 370w by the end. This is the most interesting test to me, because I got almost the same drop in temperature as my max mode test, but only used ~85% of the power. The difference may be due to different clouding on the two days (it’s a bit overcast and hazy today, I think it was only light clouds on the MAX mode day).
![Image]()
![Image]()
![Image]()
![Image]()
Improving Ecoflow Performance
The first thing to realize is that these hoses should be kept as short as possible, with as few bends as possible. The longer the hose, the greater the surface area that is available to transfer heat into the room you're trying to condition. Also, the more bends (and the tighter the bends) the more restricted the air flow is.
There's some research out there (Compression effects on pressure loss in flexible HVAC ducts (Journal Article) | OSTI.GOV) on flex ducts showing that they're pretty crappy for air flow over distances at the whole home HVAC scale (20 feet in length or so). A poorly installed flex duct can reduce your cfm by 50-75% or more! This is before taking into account bends in the ducting and is because the ridges in the duct create turbulence which restricts airflow.
Insulating the exhaust hose definitely helps, but it really depends on how long your hoses are running to the outside. Minimally, the provided hoses are one foot in length, for a total surface area of ~1.57sf. Over that amount of surface area, and with interior temperatures of 75F, the 110F air in the exhaust hose is adding about 95btu/hr of heat to the room. At the maximum hose length though (4.5 ft of hose extension and 7.07sf), that same 110F air can add ~412btu of heat to the room! That's cutting your performance by somewhere around 10% just because of the exhaust hose.
The intake hose from the outside hopefully isn't carrying 110F air, but even 85F outside air over 4.5ft of extension will add ~118btu of heat.
It's really surprising how little insulation is needed to make a big difference though. One inch of fiberglass insulation adds ~R3 which reduces the max length exhaust hose heat transfer to ~70 BTU/hr (and the intake to ~20btu). You hit diminishing returns very quickly though, R6 cuts you in half to ~35btu on the exhaust but is twice as thick of insulation.
Test Results With Ecoflow Mods
There are some inconsistencies in the test results here, notably that the baffle+manual high fan and baffle/exterior hose insulation performed worse than their non baffled counterparts. I'm not sure if that means that my baffle decreased performance (possible) or if the weather on those days effected the results. The starting temperature was significantly lower on the baffle days (~5F lower exterior temperatures), and the toshiba also performed worse on a repeat run at the end with similar lower exterior temperatures. Sun loading on the different days could also have effected things. I didn't take any sun/cloud measurements for any test, but pulling old weather data showed that every day was "fair" except the first Toshiba run which was "mostly cloudy" and the unmodifed sleep/unmodified manual high fan which were both "partly cloudy".
![Image]()
![Image]()
Overall thoughts:
Background
The reason Ecoflow performs worse than its nameplate rating is because they use the ASHRAE BTU's which does not account for the heat generated by the unit being in the room (as opposed to outside with a window a/c), leakage through the ducts, and heat being radiated to the room from the ducts. The DOE recognized that portable a/c's have these limitations and started requiring manufacturers to run a new test with portable units to give an "equivalent" rating to the old test. If you look for portable a/c's you'll see two ratings now, ASHRAE and DOE SACC.
Unfortunately there's not a single way to compare the two numbers since there are so many things that effect the rating (insulation in the housing of the a/c, insulation of the ducting, single vs. dual ducts, air leakages, single vs. variable speed compressor, etc.). This page talks about the two systems and has some ranges of comparison: SACC BTU DOE vs ASHRAE in Portable AC - PICKHVAC
Using their ranges, it seems a proper de-rating can be anywhere from 50-85% to go from ASHRAE to DOE. Dual hose is the biggest differentiator, 75-83% de-rating per the website whereas single hose is 58-70%. We know the ecoflow isn't super well insulated and does have duct leakage problems, so a 75% de-rating seems appropriate, possibly even 70%.
That all boils down to the Wave 2 being equivalent to an ~3500-4000btu window unit but using the same amount of energy as a 5100 btu window unit. If you're not using both hoses it could be as bad as a 2500btu window unit.
What can we learn to improve cooling performance?
- Run the ecoflow outside of the space you're cooling if able (duct the conditioned air into a tent or other enclosed space)
- Seal the duct connections so that exterior air doesn't leak into your conditioned space
- If you can't run the unit outside, enclose it in an insulated box to minimize heat transfer into your conditioned space
- insulate your condenser ducts (the ones that go to the outside)
- If you can duct your conditioned air away from the intake of the ecoflow it will increase cooling performance (so that cold air isn't being sent back to be cooled again). The warmest air is sent to be cooled.
- exhaust ducts should be placed further apart than the provided attachment so that hot air isn't being sucked back in to the inlet.
- If exhaust ducts can't be separated, a baffle should be added to discourage air from being recirculated.
Test Setup & Unmodified Results
Over five days I compared the wave 2 and its standard modes (no modifications) against a 5000 btu Toshiba window unit that I had out in my garage. All tests were started at ~4pm and ran for one hour to try and get similar sun loadings across days. The garage door and back door were opened for ~30 minutes prior to each test to allow temperatures, thermal mass, and humidity in the garage to equalize to the exterior. The Toshiba has a slight advantage over the wave 2 for its test because I had previously built an insert with 2” of xps to match the window unit’s size. The Wave 2 only had 1” of xps and 1/8” sheets to finish closing the gap under the door. There was also some light rain ~20 minutes into the Toshiba’s test and my exterior thermometer got wet so there is some question in the final exterior temperature for that test only. The interior thermometer was unaffected. The garage is uninsulated (except for a couple panels on the door), and is ~2000 cu ft so definitely bigger than either a/c is intended for.
The Toshiba is rated for 4 amps at 115v AC (460w). I didn’t think to take power data during its test, but a separate 15 minute followup test using my delta mini showed that AC power draw was slowly ramping up to at least 400w. A longer test may have grown closer to the 460w rating. A short 5 minute test in low mode only used ~380w AC, but I don’t have any hour long cooling tests to look at temperatures in low mode.
In MAX mode, the wave 2 peaks at 470w DC. The wave 2 does do some aggressive throttling of the compressor during its operation in all modes. I believe it targets a delta temperature of ~18F between ambient and outlet temperatures in MAX mode. If the delta temp is >18F, it will ramp down the compressor, if it’s <18F it will ramp up the compressor. I didn’t record power data during the MAX mode test, but on a shorter 15 minute test MAX mode ranged from 440-470w DC. If you start getting closer to your set point I bet it would reduce further.
In ECO mode, the wave 2 cycles through a number of “modes” to save power. It starts with the compressor on high and low fan speed for 5 minutes. It ramps up over those 5 minutes to ~380w dc power draw. Then it switches to a low compressor and low fan speed for 15 minutes. In this mode it draws ~200w. Now comes the bad part, it runs on FAN ONLY for 10 minutes, only using ~10w. The cycle then repeats. I suspect that the same compressor limiting based on delta T could reduce the 5 minute high compressor numbers if you are near your set point. Overall, without compressor limiting, the ecoflow averages ~150w DC.
The issue with ECO mode is that the residual water in the base of the unit is re-added to the air during the fan only portions of the cycle. This is BAD for overall comfort. The RH in my garage actually INCREASED during the test which should definitely not be happening while an air conditioner is running. Perhaps a longer test would get to a better steady state and start to reduce humidity.
In sleep mode the aggressive throttling of the compressor actually made a difference in power consumption. The wave 2 targets a 20F (11C) delta t in this mode, and with the fan on such a low speed there’s significant mixing of the outlet/inlet air streams (and not enough mixing with the room air) so it ends up throttling the compressor down quite a bit. This has a significant effect on overall cooling ability of the room. At the start of the test the wave 2 was using ~250w DC, but within 20 minutes it dropped down to ~200w for nearly the remainder. Total energy use ~220w DC for the hour. I did run this test for an additional hour and room temperatures didn’t drop much further, but humidity levels did continue to drop (~1.5% RH/hr).
The last test, unmodified manual mode with high fan. The compressor throttling still occurred, targeting ~17.5F (~10C) delta t. I saw a max of 400w DC near the beginning of the test, dropping to 370w by the end. This is the most interesting test to me, because I got almost the same drop in temperature as my max mode test, but only used ~85% of the power. The difference may be due to different clouding on the two days (it’s a bit overcast and hazy today, I think it was only light clouds on the MAX mode day).
Improving Ecoflow Performance
The first thing to realize is that these hoses should be kept as short as possible, with as few bends as possible. The longer the hose, the greater the surface area that is available to transfer heat into the room you're trying to condition. Also, the more bends (and the tighter the bends) the more restricted the air flow is.
There's some research out there (Compression effects on pressure loss in flexible HVAC ducts (Journal Article) | OSTI.GOV) on flex ducts showing that they're pretty crappy for air flow over distances at the whole home HVAC scale (20 feet in length or so). A poorly installed flex duct can reduce your cfm by 50-75% or more! This is before taking into account bends in the ducting and is because the ridges in the duct create turbulence which restricts airflow.
Insulating the exhaust hose definitely helps, but it really depends on how long your hoses are running to the outside. Minimally, the provided hoses are one foot in length, for a total surface area of ~1.57sf. Over that amount of surface area, and with interior temperatures of 75F, the 110F air in the exhaust hose is adding about 95btu/hr of heat to the room. At the maximum hose length though (4.5 ft of hose extension and 7.07sf), that same 110F air can add ~412btu of heat to the room! That's cutting your performance by somewhere around 10% just because of the exhaust hose.
The intake hose from the outside hopefully isn't carrying 110F air, but even 85F outside air over 4.5ft of extension will add ~118btu of heat.
It's really surprising how little insulation is needed to make a big difference though. One inch of fiberglass insulation adds ~R3 which reduces the max length exhaust hose heat transfer to ~70 BTU/hr (and the intake to ~20btu). You hit diminishing returns very quickly though, R6 cuts you in half to ~35btu on the exhaust but is twice as thick of insulation.
Test Results With Ecoflow Mods
There are some inconsistencies in the test results here, notably that the baffle+manual high fan and baffle/exterior hose insulation performed worse than their non baffled counterparts. I'm not sure if that means that my baffle decreased performance (possible) or if the weather on those days effected the results. The starting temperature was significantly lower on the baffle days (~5F lower exterior temperatures), and the toshiba also performed worse on a repeat run at the end with similar lower exterior temperatures. Sun loading on the different days could also have effected things. I didn't take any sun/cloud measurements for any test, but pulling old weather data showed that every day was "fair" except the first Toshiba run which was "mostly cloudy" and the unmodifed sleep/unmodified manual high fan which were both "partly cloudy".
Overall thoughts:
- I think ecoflow could eek more performance out of the wave 2 with some firmware updates. An advanced user mode that lets you set the target delta t between inlet/outlet would be awesome! You would use more power this way, but I think you’d gain performance. There may be some warranty killing hacks that could accomplish this.
- Unmodified, the wave 2 is coming up short compared to a standard window AC.
- I thought energy use might be better for the Wave 2 but was shocked at just how little draw my little window unit had. For some reason I thought it needed significantly more energy. Even accounting for inverter losses of 10%, you’re roughly even in power draw with the Wave 2 in MAX Mode but cooling better and at 1/5-1/10 the cost.
- My original de-rating estimate of 70-75% for the ASHRAE to DOE SACC conversion on the wave 2 was right on the money. The Wave 2 performed at 71% of the capability of the window unit in MAX mode and used similar amounts of power after adding in inverter losses for the window unit.
- One benefit of the wave 2 is how low it can throttle itself down and run its compressor. If you are in a well insulated or very small volume environment, or just trying to keep humidity levels down compared to exterior, it does seem like the wave 2 may be able to cool/heat for surprisingly low amounts of power (200w DC minimally). That’s better than any window unit that I know of.
- Portability and adaptability to a wide variety of use cases is definitely better with the Wave 2 than a window unit.
- ECO mode is worthless unless you’re in a dry desert environment without any humidity.
- 200w DC seems to be the lower limit for the Wave 2 with the compressor running and on low fan speed.
- Fan power draw appears to be ~10w per increment (10/20/30w DC for low/med/high)
- Unmodified, I don’t see any reason to run the wave 2 in any mode but MAX or manual/high fan. I’ll have to test some more, but I suspect that any mode will throttle the compressor down as that same delta t is reached between inlet and outlet (or your set point). If that’s the case, you’re going to want the fan on high to mix the conditioned air with the room air as quickly as possible. Eco and sleep modes force you into low fan speed.
