If you have a Combi boiler, or SunAmp, or pretty much any device with a built in Plate Heat Exchanger (PHE) and live in any region which has hard water (about 80% the UK population), then you will need a water softener if you want any decent life out of your plumbing installation. As far as I can see you are down to one of two options for a direct plumbing solution: the UK Harvey twin tank system and the US Kinetico range. All of the rest are niche suppliers, IMO. The Harvey system seems to be viewed as the best in terms of performance and running cost; and a few of the others on the forum use one and have recommended it, so we went with this choice.
The basic Harveys internals are boxed and rebadged by a number of suppliers: Harveys, TwinTec, Fountain Softeners, but the versions currently shipping are one of four standard Harveys models under the hood: 500, 750, 1000, 1400 (which relates to the volume of water in litres per flush). And these have a guideline maximum PPM of 320K/<model no>, so the 750 is advised for a maximum of ~ 430 ppm water hardness. This is the one that we have. Each flush on the Harvey is 17l, so the 750 flushes 17/750 = 2¼% of the water (which incidentally is under the 4% threshold on the Water Usage calculation required for Part G approval).
So in our household we use ~250l /day and so will do 10 flushes / month @300g salt per flush or 36 Kg of salt p.a. 12 × 8 Kg block salt costs £72, so our expected annual salt cost is approximately £27. A lot cheaper than some alternatives, and this has to be offset against reductions in soap and other consumables.
Now to my big bitch. I was researching the performance data of the Harveys Softene and trying to find simple performance data on the typical pressure drop vs flow. Talk about shifty and evasive. You can’t find this anywhere on the Harveys site; ditto the TwinTec, and the Fountain Softeners site. I also tried an email to my named contact at the last, but got no reply one this. There’s lots of qualitative hand-waving in YouTube videos and promotional material, including from Mr Harvey himself — all to the effect that the pressure drop is not that bad / a lot better than the competitors, etc. Calculating this pressure drop vs flow graph isn’t difficult. These curves all approximate well to a power curve, so the tester will normally takes a few PD : flow measurements over the working range of the appliance, plot them on a loglog plot; fit a straight line; and then invert this back to the power curve. OK, in the case of a water softener, the actual value might vary over the flush cycle so you might need to put in a upper / lower band on this curve, but all of this naaa-di-naa-daa sounds like they are hiding something to me.
So after a bit of Google searching I happened across a link to a PDF on a website run by a landlady in Reading (how Google found this, I have no idea). This was the standard 4 page installation guide that is most of the sites, but plus a fifth “Technical Data” page: which includes this plot. What this shows is that at a 30 l/min flow rate you are loosing 1 bar through the softener, This equates to a couple of showers going at the same time as the kitchen sink. However if you have power showers or the like, then by 45 l /min you are at 2 bar and with the other system loses, your water system will begin to struggle if you only have a 3 bar supply.
Of course if you have an accumulator, then this will be positioned on the house side of the filter, so this will greatly mitigate this limit.