Also discovered that my solar geyser often has the left (east) inlet side hot while the outlet (west) is cold. Busy building an IoT gadget to measure all the inlet/outlet/panel temperatures, mostly out of curiosity in my home assistant. Think I have figured the reasons for this... all geyser installations I have seen including my own is optimal for Northern Hemisphere. This is applicable to high pressure thermosyphon manifolded systems only, such as in this thread. Early morning, when there is no flow, the sun hits the eastern tube(s) and heats it up having a full sunrise exposed side. The thermosyphon pressure starts pushing up the water and the inlet is the closest. Flow starts in the wrong direction pulling in cold water from the outlet side. Once the flow is established, which significantly impacts the transfer efficiencies, it requires the entire panel to heat up quite a bit for the "true" thermosyphon pressure to correct the flow direction. On a colder/less radiant day this never corrects. This messes up the relationship between radiation intensity and geyser performance. Below some threshold, the heating efficiency drops off non-linear because the flow is limited being in the wrong direction and could stall for some time. Here are some ideas to mitigate the issue especially during colder days: Lifted the western side of the panel manifold slightly by placing a brick under the frame at the top (there was a brick available close by hehehe). This slightly tilts the manifold up to assist the thermosyphon from east (inlet) to west. Rotate the entire panel a few degrees anticlockwise. On a slanted roof like mine, this increased the tilt angle of the manifold (east - west) as per point 1 some more further improving thermosyphon effect. This is bit difficult in my case. Having a perfectly straight (horisontal manifold) installation in this case is detrimental for performance. Cover the east most tube left side to prevent the early morning sun heating it more than the others. Only partially to allow it still getting full sun later on during the day. This should lower the chances of it pre-heating the inlet port, causing reverse flow in the morning. A circulation pump to force the correct flow will eliminate and significantly improve colder day performance. Did buy a pump for other reasons and first want to see improvement results from above fixes. Think this happens to be the reason why GW is so successful to improve geyser performance by ensuring correct flow direction especially during colder days using their circulation pump setup. On a new installation, rotate the entire hot water tank 180° with the inlet on the western side and adapt piping accordingly... southern "hemispherise" the installation. This will solve mostly the east rising sun issue on the inlet side. Insulated the inlet and outlet lines between the panel and hot water tank. A lot of heat is lost during the day as the hot water circulates while exposed to the colder environment. Trust this will help improve your system(s).