Rocketstove water heater efficiency

Anybody knows something about the efficiencies of their rocket water heater? So we can compare. I build one with a barrel over a barrel system (so the heat riser is insulated) with thermo siphon working to heat water for a hottub. I calculated an efficiency of about 25%. This calculated with 15MJ per kg wood against the temperature rise of the water being 960 liters 6degC to 40degC. The low efficiency (imo) could be due to the heat losses of the bath (which I covered during heating) and the draw of the system not being perfect. Since I have sometimes a bit smoke and flames from the burning chamber (back draft). I lead additional the exhaust under the bath through a concrete tunnel. The exit exhaust temperature I measured is around 30 degC and the "smoke" i see is mostly water vapour (i concluded since it does not have the smoke smell of an ordinary fire). I measured CO with a safety gas sensor, which measured over range so above 1000ppm. Since the rocket heater is located outside the heater is a bit humid of course also the ground where heater is located is humid, and reduces the efficiency further. The system is a 8" system.

A picture to show what i built

I used about 32kg of wood to heat the 960liter water from 6degC to 40degC after 2 hours the water temperature was still 36 degC

It took 4 hours to heat to 40degC . After the 4 hours it took 2 hours to cool down from 40 to 36. So there is a lot of latent heat in the system

And in operation

Is there an actual chimney to vent the carbon monoxide away from the bath?  

No there is chimney what so ever. And it is one of things I am thinking about. It could also increase the draw. The CO disperses however quite quickly I measured with my CO meter. (Since we are outside) Further i like to do more measurements when using less wood. So if it is a  problem I am not so sure.  And please note when in the bath on temperature no more wood is added, just the latent heat is sufficient for the bath to stay on temperature.

Two questions come to mind:
1) what are the temps of the outside of the barrel, the pipe going into the wall, the pipe coming out of the wall?
2) you mention a concrete tunnel under the bath; what is it made of and how is it insulated from the surrounding soil?

And one comment: Those pipes don't appear to be insulated. That would have to be first on my list of modifications.

The temperature outside the barrel depends on the time during heating and on the location of the barrel, and is only a bit higher than  the  highest water temperature of the bath. Which is about 45degC on the top at the end of the heating. Lower on the barrel the temp is lower. (Mostly depending the supply temperature). Due to these temperatures I concluded insulation would not benefit the efficiency much.  

The bath is constructed of concrete elements, which were originally for plants intended (on the first picture you can see the particular shape of the elements.) The bath is not insulated from the soil other than the air in the concrete elements, which I thought to be sufficient insulation, the exhaust gasses leading under the bath I saw more as an additional benefit. Thinking it would cause a small efficiency increase, the main work is done by the barrel and the pipes. I realise the insulation is not ideal however after 24 hours the water is still about 25 degC at an outside temperature of 8degC.

What I hope to learn from this tread is how much wood people need to heat their hottub for a certain temperature increase for a certain amount of water.