All "Secret Document" incidents = Roughly 1 or 2 per cent of a single luminous wristwatch dial falling in the sea
Incident 1 - August 2007
HMS Superb discharges some ordinary demineralised water into the loch through a pipe which had previously had water with tritium in it.
Subsequent investigation determines that as much as 24 MegaBecquerels of tritium could have been released directly to the loch, rather than being released later from the shore pipe as normal. This equates to a quarter of one percent of the amount that could be released by dropping a luminous watch in the loch. Hundreds of times this amount of tritium is put into the loch every month on average, with full SEPA approval.
Incident 2 - February 2008
HMS Torbay pumps out ordinary demineralised water into a tank barge which already has some water containing tritium in it. Because of a fault with the barge level gauges, the barge briefly spills water into the loch through overflow pipes as its tanks fill up. The pumps are turned off straight away after the overflow begins.
Subsequent investigation determines that the "amount of water spilled was in the order of litres, not tens of litres" and that the maximum possible tritium spill was less than 100 MegaBecquerels (less than one hundredth of the amount one would release by dropping a cheap luminous watch in the loch). Dozens of such incidents would have to occur every single day for a year before Faslane breached its SEPA tritium limit. All the water would have subsequently been put into the loch anyway a couple of hundred metres away complete with tritium, with full SEPA approval.
Incident 3 - 2004
Another incident of this sort occurring with HMS Trafalgar is alluded to in the released SEPA material, but not described in detail.
There was also potential for sub-microscopic amounts of Cobalt-60 to be released in these incidents: this isotope gets generated in submarine pipework and can sometimes be detected in the tritium-y water that comes out of them. In every case described here, however, cobalt levels were either undetectably low or negligible compared to those of tritium - itself only present, as we have seen, in insignificantly small amounts.
So to sum up. These "liquid radioactive waste leaks" were utterly, mindbogglingly unimportant. You could go out and buy a cheap, fully legal luminous watch, throw it into the Gareloch, and you would have released many, many times more radioactive contaminants than the Navy did here. You can wear that watch, and if it's a cheaply-made one (eg an older Swatch) the tritium will leak out as though you had a small submarine base, operating just one or two subs, strapped to your wrist. Tritium will (aaiee!) get absorbed into your body, emerging in due course at detectable levels in your urine. But you'll be fine.
And yet each of these naval "incidents", less significant than a wristwatch accident, is treated as a big deal by the authorities. Reams of paperwork is generated, both in the Navy and at SEPA. Handling of water with tritium in it - which everyone concerned admits is low-level waste, far less concentrated and dangerous than luminous paint - is upgraded to the highest level of nuclear procedure at Faslane as a result, generating mountains more paperwork, records, improvements and modifications to the base, new barges etc. SEPA issues fatuous statements that - if Faslane were under its jurisdiction, which it is not - the base might be subject to "regulatory action", or even shut down.
That, quite frankly, is insane - but that's the freakish world of nuclear safety for you. This is the reason why nuclear power didn't turn out to be too cheap to meter: because everything nuclear or even just radioactive has been regulated almost to the point of impossibility, as a result of hysterical, terrified ignorance on the part of politicians and the general public.