Disastrous tunnelling delays underline folly of Snowy 2.0 pumped hydro scheme

The latest revelation from the hapless Snowy 2.0 pumped hydro battery project is the staggeringly slow progress of all three tunnel boring machines (TBMs), resulting in extensive delays and cost blowouts.

This news hasn’t come from Snowy Hydro, which continues to issue positive updates and unrealistic tunnelling estimates, but from independent investigations.

The TBMs are massive machines – 11 metres in diameter, 140 to 200 metres long, weighing over 2,000 tonnes – boring 33 kilometres of tunnels, but extremely slowly as we have come to discover.

Named after three notable women – Lady Eileen Hudson, Kirsten Banks and Florence Violet McKenzie – their boring has been anything but notable.  All are years behind schedule.

Focussing first on TBM Florence.

She was commissioned with much fanfare on 22 March 2022 (though no mention was made that the commissioning was six months late). 

But of far more concern is that over the past year Florence has only bored 200 metres, barely more than her length.  (Her tail is just inside the tunnel portal.)

She has averaged 60 centimetres a day!  

If she had been operating at her design specification of 30 to 50 metres per day she would have bored 200 metres in less than a week and would now be at least nine kilometres into the seventeen kilometre headrace tunnel from Tantangara Reservoir to the future Marica surge shaft (i.e. be over halfway).

Having bored just 1% of the headrace tunnel, Florence has no chance of achieving the September 2023 schedule.

At her current ‘speed’ Florence will take another 70 years.

This astounding and highly embarrassing revelation has only become evident from delving into a seemingly innocuous ‘Snowy 2.0 Headrace Tunnel Update’ issued on 16 December 2022 advising that “Snowy 2.0 teams are monitoring a surface depression that has emerged above Florence”. 

The update referred to “the ground conditions encountered by the TBMs on Snowy 2.0 being highly variable in fluctuating groundwater conditions”.

The update gave the assurance that “tunnelling is continuing while work to remediate the surface depression is carried out”. 

However, we have heard that is not the case and that Florence has been immobilised for the past two months and the surface depression has yet to be remediated.  We have even heard rumours that Florence is irretrievably stuck, unable to go forwards or back.

The update gave no hint of the magnitude of Florence’s problems or her miniscule progress since commissioning. 

An intrepid bushwalking colleague went searching for the surface depression, starting several kilometres from Tantangara where Florence was expected to be. 

He ultimately found a 15m wide x 4m deep sinkhole, a tad more substantial than a ‘depression’, about 175 metres from the portal.  (We have rounded Florence’s progress up to 200 metres in case the update is true and she has continued tunnelling since mid-December.)

Whether or not Florence is currently boring, she must surely have set a new world record for slow speed tunnelling.

No doubt her dire predicament has been the subject of frantic investigations over the past year, though unbeknown to the public.

We trust that the emergence of the sinkhole will have widened those investigations to include the possibility of more damage than was previously expected to Kosciuszko National Park and its water catchments.

The EIS for Snowy 2.0 acknowledged that tunnelling will have surface and ground water impacts, but considered them to be insignificant even though streamflow drawdowns of up to 50 metres were predicted in some areas. 

There was no mention of the possibility of sinkholes or their potential for additional impacts on Snowy catchments.

The only other public update concerning Florence since her commissioning was in October 2022 after she had been stuck/stopped for an undisclosed period, rumoured to be a couple of months. 

That positive update, ‘TBM Florence on the Move’, announced that “TBM Florence continues excavating the headrace tunnel following successful reinforcement works … having encountered highly variable ground conditions and high groundwater inflows, as anticipated”, with an assurance that “Florence has been specifically designed to encounter these variable ground conditions”. 

Again, just a minor, expected issue, but now it’s full speed ahead, with no mention of the extent of Florence’s problems or lack of progress.  

The same positive spin was repeated at a Senate Estimates Hearing on 7 November 2022 when the Acting CEO was questioned about Florence being bogged. 

He responded that “the TBM hasn’t been bogged … we have been progressing the TBM through that expected [variable ground condition and high groundwater inflow] zone with care and diligence … that is being worked through systematically … the machine has been paused at various stages … it is not right to say that it has been bogged”.  

He just overlooked informing the Senators that Florence had been effectively ‘paused’ ever since her commissioning, seven months previous.

Whilst Florence is the standout laggard, the two other TBMs have also failed to tunnel at anything like their specified capability.

To give some background, the following diagram from the Snowy 2.0 Main Works EIS shows the respective tunnels being bored by the three TBMs (note that references to MAT and ECVT should be reversed).

TBM Florence is boring the 17km headrace tunnel (blue line in diagram), but is ‘paused’ at Tantangara.

TBM Eileen (ochre) completed the 3km main access tunnel (MAT) from the surface at Lobs Hole into the location of the cavern for the underground power station on 11 October 2022.  However, she was two years late and took three times longer than expected.

According to the Snowy 2.0 EIS, tunnelling was to commence in April 2020.  Subsequently, at a Senate Estimates Hearing on 20 October 2020 the CEO advised that first operations would be at the end of December 2020 and tunnelling to the cavern location would be completed five months later, by May 2021. 

Ultimately, Eileen was commissioned in June 2021, fourteen months late, and then took sixteen months to complete the journey.

Eileen is now being dismantled and relocated to bore the 8 km tailrace tunnel from Talbingo Reservoir back to the cavern.

TBM Kirsten (red) was commissioned on 6 December 2021 (nine months late) and completed the 3km emergency egress, cabling, and ventilation tunnel (ECVT) from Lobs Hole into the cavern location in twelve months – faster than Eileen but taking twice as long as expected. 

Kirsten is continuing to the 2km inclined pressure shaft to Marica to meet up with Florence (if she makes it).

So, the two 3 km-long access tunnels into the station cavern have been completed, albeit years late.  Now there are 27 kilometres of water tunnels to be excavated – the headrace tunnel (Florence), tailrace tunnel (Eileen), and pressure shaft (Kirsten).

Unless all three TBMs miraculously speed up to their design specifications there will be further substantial slippages in the construction schedule in addition to the two-year delay so far.  

Florence’s predicament is the most problematic, particularly as she has the longest distance to excavate and has been paused at the start for nearly a year.  

Even if Florence was un-paused and immediately attained Kirsten’s speed, the ‘fastest’ of the three TBMs, she would take six years (i.e. 2029). 

Yet Snowy Hydro continues to publish wildly over-estimated expectations.  

For example, the latest Snowy Hydro Corporate Plan 2023-2027, issued on 28 October 2022, sets an astonishingly optimistic “key delivery milestone for completion of all three TBM drives by 2023”.  This is not remotely possible.  Nor is the Corporate Plan’s estimate of first power in 2025/26!

The tunnelling delays are starkly evident by the chock-a-block stockpiles of tens of thousands of tunnel lining concrete segments at the Cooma factory and elsewhere, that would have been installed if the TBMs had been performing to specification.  The storage area at the factory has already been expanded.

The factory continues to produce segments 24/7, but none of the TBMs are boring at present.

Not only are the delays extending the construction schedule, but they will also be blowing out the cost of the project even further.  No doubt the extra costs are the subject of intensive negotiations between Snowy Hydro and Webuild, the main contractor since Clough went broke (with Snowy Hydro being its largest creditor).  

We are aware that Webuild submitted overrun claims a year ago totalling $2.2 billion.  Additional claims will have been submitted since and even more will be forthcoming in the remaining six or so years of construction. 

The tunnelling imbroglio raises many questions:

• When will Snowy Hydro start providing un-exaggerated and fulsome information on Snowy 2.0’s progress and realistic estimates?  
• Why have none of the TBMs performed anywhere near their design specification?  
• If “Florence has been specifically designed to encounter variable ground conditions”, what went wrong – unidentified ground conditions, bad design/construction, poor operation?  Or was it simply due to choosing an unsuitable dive site?
• What has been tried to ‘fix’ Florence and why is it taking so long?
• More importantly, can Florence’s predicament be resolved:
  • if so, how long will it take, how long before she completes the headrace tunnel, and how much extra will it cost
  • if not, what is Plan B?
• What additional precautions are proposed to mitigate environmental damage to the Park, especially its water catchments?  
• Has the Government been kept informed of the tunnelling problems over the past couple of years and, if so, what has been its response?
• When will Snowy Hydro provide a realistic date for Snowy 2.0’s completion?  The estimate of first power in 2025 (website) is clearly out by years, as is the latest advice to AEMO of full commercial use by December 2027
• Finally, what is the impact on Snowy 2.0’s capital cost, still stated as $3.8 – $4.5 billion?

Well before the extent of tunnelling delays have now become evident, energy experts predicted that Snowy 2.0 would take at least twelve years to complete (rather than four as initially announced), at an all-up cost including transmission of more than $20 billion (rather than the $2 billion announced), and require far more than the $1.4 billion of public funds already provided (rather than zero as promised by Snowy Hydro).

That $20+ billion will ultimately be paid from the public purse and electricity consumers.

Snowy 2.0 never stacked up – not economically, not technically, not environmentally.  Its eye-watering cost blowouts and massive impacts on Kosciuszko National Park cannot be justified for providing occasional long-duration storage.  Snowy 2.0 will not provide a net benefit and will never pay for itself.

Surely these latest tunnelling revelations and the continuing provision of unrealistic estimates by Snowy Hydro, on top of the numerous other failings of the project, will impel the Federal Government to commission a comprehensive independent review.  

Such a review has been urged for years by energy experts in order to obtain an accurate assessment of the project (e.g. here, here and here).  

Snowy 2.0 cannot be allowed to continue blundering on as it has for the past six years, regardless of the consequences.  Public monies cannot continue to be squandered.  A few billion dollars has been sunk into the project but there are many more billions to be saved.

A least-worst pathway needs to be urgently developed to address this debacle.

Ted Woodley is a former MD of PowerNet, GasNet, EnergyAustralia, China Light & Power Systems (Hong Kong); Executive Member of the National Parks Association of NSW