It is not at all clear what efforts were made to control the Troggy during the building of the Severn Tunnel. The chief Contractor, Thomas Walker noted in his book the following*...

Sir John Hawkshaw had further decided that, in order to reduce, if possible, the quantity of water to be dealt with — which, we knew, to a very considerable extent came from the loose bed of the little river Neddern — to construct a concrete invert in the bed of the brook for a length of nearly 4 miles.
This work was undertaken in August [1884?], and was completed by the 7th October...

Quite where the invert was built is hard to tell. The sensible place would probably be where the water disappeared into the ground at Llanvair Discoed, or possibly over the Great Spring at Portskewett, however, there is no sign from the map of anything other than natural channels and, indeed, there are still plenty of sinks at Llanvair. The only section that looks artificial is between the Roman road west of Caerwent and the Upper Whirly Pool site (through there are short stretches of straight channel at Caldicot Pill).

If one looks at the 1887 Map, it is clear that the dotted boundary follows the Troggy to the east of Caerwent, however, below Caerwent the boundary undulates as if following a natural channel and matches the contours, while the stream is incredibly straight (though only for ~1 mile). Interestingly, the boundary by the Upper Whirly Pool leaves the Troggy's current course, which runs at the foot of a bank, and heads across to the Whirly Pool, perhaps indicating an older channel position. However, this would have been abandoned long before the Tunnel was built, as the channels are as shown in this manner on the 1775 Map.

Indeed, on the 1775 map the boundary leaves the stream altogether and heads upslope to Slough Farm. More notably, the 1887 Map shows the boundary following the Troggy at the Lower Whirly Pool Site with its new neighbouring drainage, whereas the 1775 map shows the boundary moving to the top of the bank around the, then undrained, area (though the drainage hardly seems to have worked as the area is still very flooded).

Local knowledge also has it that the Whirly Pools were filled during the Severn Tunnel construction*. The 1887 Map appears to show various pools, drains and weirs around the Upper Whirly Pool which may be to do with the Tunnel construction, or may have been put in place any time since 1775.

However, sealing the resurgences seems a little counter-intuitive, and may not have been necessary - or even attempted - Walker notes...

The whole of the ground in the marsh is rotten, and before the tunnel was commenced there were enormous springs of bright clear water rising up in several places...

When the tunnel was being made, and [the Great Spring fissure] was unfortunately tapped in the rock between Sudbrook camp and Portskewett village, all these underground channels poured their water into the tunnel itself, and almost every well and spring, and the little river [Troggy] itself for a distance of more than 5 miles from the tunnel, became dry.

Subsequent pumping of the tunnel, which continues to this day, reduced the water pressure at the tunnel (>145 feet [44.2m] below sea level) to nothing. Given Walker's report on the accidental drying of the springs before, this is almost certainly enough to keep the springs dry. That the springs haven't been entirely capped is suggested by reports that some were open in the 1950s (See Note), that 1:2500 OS maps show the stream leading from the site of the Upper Whirly Pool site as active, and that the lower Whirly Pool site may be producing enough water to clear plantlife (See Visit Details),