South Sister St. Marys, Tasmania

South Sister Evidence

risks to water supplies

Comments on Proofs of Evidence by Experts for Forestry Tasmania
- re Coupe NI 114A, South Sister, St Marys.

D.E. Leaman August 28, 2005

Dr E. O'Loughlin

Dr O'Loughlin reviews the yield scenarios of Lane and claims that, because the trees to be harvested are largely aged 30-55 years, that Lane has over-estimated future yield declines. He argues that it will be near zero. He further argues that because overall yield changes will be small effects on springs will be negligible.

I totally disagree with this point, a critical point. There are several grounds for disagreement.

  1. We do not know the actual rainfall across the coupe or mountain. Everyone assumes a value or infers it, but the value is critical to all calculations.
  2. I have seen no evidence which convinces me that the tree age distribution is as claimed. Is this new data?
    The Forest Practices Plan suggests the older trees are to be used (based on size specifications).
    Why did Lane, who was independent of the present argument, believe the trees were older than 55 years?
    The 1946 photography indicates a reasonable stand of trees at that time. (Note that photo comparisons must allow for scale of photo and sun angle, as well as canopy) I would argue, pending a valid statistical dendrochronology of the forest that the general age of substantial trees is likely to be 120-150 years on the basis that the first miners - as was the custom of the day - essentially burnt and cleared the original forest (1845-1885).
  3. Previous reports by Roberts, McIntosh and Lane have offered various estimates of short term run off after clearing and long term reduction of flow during growth. These estimates are 60-83, 114-156 and 50 ML/ha/year respectively for the relevant (?) rainfall range. Further, O'Shaughnessy (attachment to Erskine) suggests that the yield change is unlikely to exceed Lane's estimate but he does note that quite a few trees are more than 100 years old.

None of the above authors have used data local to the Break O'Day catchment of South Sister and so all conclusions, including Dr O'Loughlin's, must be estimates.

We cannot allow such estimates to be regarded as anything more than indicative. This is not a criticism of the theory and research behind such estimates.

Indeed, my own continuing research on East Tasmanian catchments (one of which is the Break O'Day above Killymoon and thus includes South Sister) has led to a local modification of the Kuczera curve which, with due allowance for all the factors noted by Lane and Erskine, led me to suspect that the loss might be 50-100ML/ha/yr depending on rainfall and at the top end if the trees culled are actually more than 100 years old.

I did not present this data, any theoretical outline, my modified curve (although referenced - #30, 58) or any detailed argument for it because I did not need to: other authors covered my range from the point of theory.
My estimate, however is derived from actual observations of forest behaviour in eastern Tasmania.

Thus, Dr O'Loughlin's conclusion might be valid, or close to reality, if - and only if - the trees are young (less than 55 years).
This must be established.


I agree with Dr O'Loughlin (and others) that the soils are permeable and have low erodibility in many situations.

But I disagree with the erodibility concept and its hydrological implications because coarse particle movement in soil is not the issue.
Removal of fines, and their transfer into the fractures of the underlying rock, is. Such fines may not block soils, form horizons, or affect soil water movement - but they may block - in an unpredictable and variable way - fractures.

An introductory quote from Rock fractures and fluid flow - contemporary understanding and applications by US National Research council 1996, says it all:

'Suspended solids or liquids that solidify may be added to a system deliberately - for grouting of propping of hydraulically induced fractures - or inadvertently during injection of water for aquifer recharge, aqueous waste disposal, well drilling and completion, and water flooding. The introduction of solids into a fractured rock mass will usually decrease its bulk permeability. This permeability decrease is one of the desired results for grouting but is generally undesirable in cases of inadvertent injection of suspended solids.' page 430.

The particles referred to are clays, colloids or silt sizes.
There is a large engineering and petroleum literature on this topic.
I quoted only a summary reference in my evidence.

Dr O'Loughlin does not distinguish subsurface conditions which include soil AND rock. Most of the water storage and transfer occurs in the rock (which everyone else terms deep circulation or drainage). It does not occur in the soils.

A clear distinction must be drawn between soils and rocks. The subsurface must never be described just as 'soils', for it is not. This is an engineering convenience and reflects the common lack of appreciation by surface water hydrologists.

Further, the soils (and slope deposits) are not uniform in properties, composition or thickness and consequently water passing through them may slowly pipe and erode them - and the fines must go somewhere: the only place is into the fracture system below.
A fracture system can be modified hydrologically in this way and the effect may be temporary or permanent. Ref. 31 in my evidence provides an example of a water bore which failed due to fracture blockage.


Dr O'Loughlin (and others, below) have not appreciated the unusual sitting and yields of the St Marys bore in Tasmanian terms. Only Weldon understands the issue of exceptional location and the fracture system.
I therefore disagree with any comments made by him, or Latinovic whom he quotes.


Dr O'Loughlin's comments on spring supply is based on assumptions about soil and water connectivity. He mentions perched supplies but these rarely act permanently or generously and we are dealing with springs which are perennial but seasonal. He further argues that small changes in catchment behaviour must mean small changes in head and small changes in spring flow.

I disagree with this because it ignores the erratic nature of springs.
Springs may have some soil pathways which link to rock fracture pathways - and storage which guarantees supply - but disturb any of these and piezometric head differences become irrelevant.
The observed changes near recent roading on the coupe demonstrate just how sensitive such systems are: new springs arise and others reduce. None of this was considered or predicted: but it happened.

Long experience with farm supplies and many queries about spring management has taught me that the best course is not to interfere in any way with a spring if you wish to continue using it.

Risk is not small, just because some changes are small. The entire system must be understood and, above all, one needs to know just where - and by what means - the water is supplied. A spring on Dublin Town Road may be fed from fractures with a recharge on the south side of the mountain.
This is simply not known and it is unsatisfactory to claim little risk, without such knowledge.


Dr O'Loughlin suggests that stability could be decreased by activities but in the absence of data about the geometry of old slides, the materials, their thickness, their properties, groundwater levels, it is simply not possible to say.
I agree with this judgment.
Observations are required on these elements.
My claim has always been the interplay between old failure surfaces, the groundwater system and its control of springs and lows, and the possibility of additional minor failure. The critical issue is control of the water system.


Only O'Loughlin appreciates the risks which might be involved in actions on the coupe but he does not rate them properly.

It is worth considering the RISK EQUATION here since virtually no one else is considering it.

Risk = probability of an event happening x damage done if it does.

In terms of Dr O'Loughlin's evidence there is a good chance that water yields and storage will decrease (unless tree ages can be proven), or that disturbance will change at least some springs, or that disturbance will modify some fracture flow.

The damage done: the communal spring may be reduced or terminated. Or it may be unaffected. If it is damaged then the issue is how is the water supply replaced? Consequently the damage done could be high.

Thus, even if the chance of the event is low - as he claims, the risk is high.



Dr Erskine offers a detailed outline of forest theory and water use. Contrary to his assertions I agree with it (see above) and, indeed, have refined the pattern of run off and subsequent draw for local catchments and generated a 'modified Kuczera curve' for those catchments.
I can, however, find no reference to any Tasmanian research in his outline. I quoted four in my evidence.
Any conclusion he draws must therefore be tempered with the need to allow for local conditions (as he notes on a Forestry Tasmania website reviewing a report on Blue Tier).

The issue, as with Dr O'Loughlin, is the age of the trees and what ERROR BAND we place on values used for calculation. In fact, both Erskine and O'Loughlin fail to offer any range which might allow safe or conservative operation. This is unscientific. The reason? No one has any data actually relevant to the coupe.

In such a condition, some caution is required.
My discussion for Dr O'Loughlin underscores this. The range in water loss could be 0 to 150 ML/ha/yr but there is general agreement that it is likely to be of the order of 50 to perhaps 100 ML/ha/yr using Lane's more careful assumptions and my own curves for the region. Again, this is why I did not need to cite all the theory.

And duty of care, given the present state of knowledge, would require us to assume something in the 50-100ML range pending further, appropriate, work. No one should claim more, and no one can justify less - unless the trees are indeed less than 55 years old.
If the forest operation uses a preponderance of older trees, or even a large proportion of them, then the upper part of this range should be used.

I would make two other comments on the theoretical discussion.
His Figures 18 is not relevant since it assumes repeated harvesting which is not proposed at South Sister.
My east Tasmanian research, and the two references (#29 my evidence) not cited by Erskine, show that the run off phase assumptions may also be invalid for this region.

Basically, no one knows the specifics for South Sister and we should not be too precise about conclusions. The real question is, will there be any collateral effects from even the small water loss range which can be accepted with some confidence: Erskine does not discuss risk.


See comments as for Dr O'Loughlin. Erskine is not familiar with Tasmanian hydrology nor the unusual particulars of this bore and quotes from general reference are quite invalid (see also Weldon). He is not in a position to comment on effects emanating from any part of the catchment.


See also comments for Dr O'Loughlin. Erskine simply assumes that everything is a soil. There is no appreciation of the fracture system, nor its relationships to the soils or the spring system.


Erskine does note that there are more slides than have been documented. This was also my belief, but one not appreciated by McIntosh. Erskine considers them inactive and they may well be. The real issue with them, is their hydrological connections and his observation supports my contention that there may be many of these.



I agree with Weldon that clays may disperse and may affect soil capacity and that working may compact soils and affect infiltration.
I disagree regarding the protection of the communal spring. He argues that the buffer zone should be adequate to protect supply but we have no idea where the water is coming from, nor its path. The proposed buffer may not be relevant.
He does apply a caveat to this effect.


I agree with Weldon on this matter, or really he agrees with me, with only one point of disagreement which I think is oversight on his part.

Only Weldon, of the experts presenting evidence for Forestry Tasmania, has any knowledge or experience of the groundwater issues which might be involved in the town supply and is the only one qualified to make comment. He therefore, disagrees with the comments and criticisms made by McIntosh.

The point of disagreement involved the Triassic basalt.
The driller's and consultant's report for the hole (SM3) shows that no water entered the hole until the basalt was encountered - even though the flow was thin (3 metres). The volume amounted to about one third of the bore yield. The basalt is clearly a factor.
His comment that the bore has an 'exceptional yield which indicates that the bore is located in a highly fractured zone known as the Cornwall Fault. The fault also passes through the flooded Jubilee Mine working.'

I agree with Weldon that we do not know what effects working South Sister will have, nor how long they might take to show up. This had also been the view of Latinovic from Mineral Resources Tasmania.

There is a lot at risk here if the wrong decision is made with respect to the town supply.



Dr McIntosh claims that the local soils have low erodibility and therefore should not affect soil hydraulics or springs.

I agree that the soils are not particularly erodible and that they lack clay bands or skins. But they do erode and turbidity increases do occur at times and fines may enter fracture systems. Residents have reported that spring flows are weather and season related and that higher spring lows are associated with lower quality. This is reality. The consistency of these observations is not known.

As discussed for Dr O'Loughlin there is a huge literature on fracture and pore sealing in consolidated rocks: the storage and transfer media at South Sister. Dr McIntosh clearly does not appreciate this and I must disagree with any conclusion he draws on the basis of soil problems.


I agree with McIntosh about the origin of most ancient or relatively recent failures in the South Sister area but I disagree about the size of some of them. I judge one to be some hundreds of metres long.

I further agree that the risk of landslide across most of the coupe is low but the hazard in this case related to the number and connection of old failure surfaces and their control of water movement and transfer of particulates. Springs depend on such factors. Working of the land may well affect such foci but since we do not know just where they are (no forest plan shows such old surfaces and Dr Erskine agrees there are many), then we are not in a position to assess potential risk.


See comments for O'Loughlin and Weldon.
I disagree totally with McIntosh on this matter.


Dr McIntosh states that wet or damp areas will not be worked and that problems cannot arise.
I disagree that this is feasible.
No Forest Practices Plan indicates where such areas are, many are not obvious and could be easily worked through, and working could create others - as has happened already. Just who supervised all this given present knowledge?

Dr McIntosh advises that buffer provisions are adequate but this presumes shallow and soil based flow only; yet he, Erskine and O'Loughlin also refer to deep circulation. Once this concept is introduced, in the absence of knowledge about fracture flow, direction of purging, and recharge, no one can possibly state that the buffers are adequate.

I further disagree with McIntosh that small flow changes are not relevant to spring behaviour but this is not long term experience. Rainfall patterns, a short drought, land disturbance - all seemingly small, can have serious consequences for both quality and permanence.


Very little evidence by anyone considers the risk equation as discussed under Dr O'Loughlin.

There is little known about the hydrology of many places, including South Sister, that we must act with due diligence when an irreplaceable resource might be hazarded.

Monitoring programs, at very least, are required before final decisions and plans can be made for the working of forest on South Sister.


I have long urged monitoring and assessment before planning and, in this case, both Mr Latinovic (MRT) and Mr Weldon (for Forestry Tasmania) are urging some increased understanding at very least.

Monitoring and evaluation must review two conditions: the nature and supply of the spring system, and the nature and supply of the town system. There is some overlap in the requirements for these studies.


  1. Basic borehole monitoring grid:200m separation on coupe area plus some sites outside the coupe. This is the master grid. Hole depth: at least 20m or well into bedrock (guaranteed, not large failed blocks).
  2. Monitor rainfall at three points on the mountain in order to assess local variations in inputs (N, centre, S side).
  3. Monitor water levels, conductivity and turbidity at least daily until rain response understood and then at least once weekly. The swap over may require several months of daily readings. Automated recording would make this easier.
  4. Relate levels to rainfall.
  5. At end of first year review patterns for inputs and seasons.
    It may be clear that some areas are anomalous (probably nearer or more distant from main arteries of flow) and a few additional bores may be needed. It may also be necessary to observe extra holes near the major spring outlets (such as Dublin Town Road in order to assess connections to failure zone).
  6. It will probably be necessary to maintain the program for at least three years. This is due to the irregular and often intense rainfall pattern which may occur in any season.
  7. Evaluate responses in terms of spring flows, levels and fracture systems.
  8. During the monitoring period some of the bores should be pump tested and the others used as observation holes in order to fully assess hydraulic properties. Tracing methods may also be required.

It should be possible to assess source areas, pathways and safe buffer requirements from such observations.


  1. In order to demonstrate any connection with South Sister at least three lines of bores must be drilled to at least 40m across the valley of Gardiners Creek (Cornwall Creek zone). No line should be less than 30-60m above the elevation of the station so that there can be no doubt of flow direction and involvement of other parts of the catchment. Holes should have a spacing of about 50m to define fault zone and to drill rocks clearly to either side. Infill holes (20m apart) must be in the fracture zone.
  2. Monitor levels and flow patterns. Add tracers at the uppermost line and at the Jubilee Mine.
  3. Pump test at least two bores/line; one line each year.
  4. Match and contrast hydraulic properties across and along the fault zone.
  5. Some additional bores are required up the valley and near the town bore although the present array near the station might suffice for conditions to north and west.
    Up valley bores: no more than one or two.
  6. The town bore array must monitor the drawdown pattern for the local system and assess whether there is any flow bias, and from which direction.

All these pieces of information should allow an assessment of the efficiency of otherwise of the fault system, whether and how long flow is directed along it to the town, and what proportion of town supply is so generated.

D. E. Leaman
August 28, 2005

Default Colours Less Contrast More contrast

4156 (1, 1, 8, 289)