The results are in, though how accurate they are?
There were 2 things I wanted to check, how long to boil for and how low (high?) the alkalinity got.
I measured my tap water at 262.5 mg/l CaCO3, then put it on to boil. I then took samples at 5,10,15,30 and 45 minutes after it reached boiling. The samples were left for 3 hours to cool and settle. I then measured them again using a salifert carbonate hardness / alkalinity test kit.
The 5 minute sample measured 68mg/l initially, it then went blue again and required another drop to bring it back to pink, making it 70 mg/l.
The 10 minute sample came in at 39.5mg/l and again returned to blue, ending up at 45mg/l.
All the other samples came in at 39.5mg/l and stayed there.
I'm assuming the reason the first 2 samples changed back to blue was that there was still some chalk in suspension which went back into solution when the test was carried out, maybe the particles had grown after subsequent boiling so had dropped out in the other samples.
Conclusion - yes you need to boil the water for 15 minutes to remove temporary hardness, boiling for longer won't reduce it any further.
Secondly I probably need a more accurate measuring method but it seems our figures are reasonably close.