Diversion: Te Mato Vai Betterways.
Compiled by: Andy Kirkwood. May 2021.
Contact: [firstName] @ islandbooth.com.
This is a working draft. Published 19 May 2021. Updated 8 Dec 2021.
Betterways is a series by Te Vai Ora Maori showcasing methods of remediating the new Te Mato Vai drinking-water treatment system for Rarotonga, Cook Islands.
Diversion means collecting only clear streamwater. During heavy rain, times when large volumes of organic matter is washed into waterways, collection will stop. Water collection will resume when the streams clear.
The new Te Mato Vai drinking-water treatment system has three waste by-products:
The operators drain the sludge from the settling tank. Sand filter cleaning cycles are automatically triggered by the build-up of fine material. The sludge is stored in onsite ponds.
Reporting on the PACl trial in December 2020, project manager GHD advised that sludge volumes were ‘far below what has been estimated’. In April 2021, To Tatou Vai is now saying there is ‘far more than was anticipated’.
A request for expressions of interest for the supply of water treatment chemicals was issued in May 2019 by To Tatou Vai. The (dry) volume of the coagulant chemical polyaluminum chloride: 75~100 tonnes (annually).
Formal assessment of the resulting sludge volumes was conducted in March 2020 by an independent engineer appointed by the court to answer landowner concerns regarding the environmental impact of the new system. The sludge estimate is based on an standard industry-calculation, the through-put of the new system (plant utilisation); and data on Rarotonga’s source water quality from the 2014 Watercare analysis.
Annual solids generation from all ten sites, assuming a plant utilisation of 60% is estimated to be 161.6 tonnes per year. At 4% [dry solids], this is equivalent to 4,040m3 of sludge. As discussed in (n), sludge at this solids content will not be spadeable and will require pumping and dewatering. This is likely to require the use of a portable dewatering rig or a sucker truck to pump out and transfer the sludge to a dewatering facility. Assuming a sucker truck capacity of 8m3, this will require a total of about 500 trips per year for harvesting sludge.
At more familiar scale, 4,000m3 is equivalent to a 20-foot shipping container for each of the ten intakes, every month; a total of 120 container loads of sludge per year.
The sucker trucks used in Rarotonga for clearing domestic septic tanks are typically 3m3; using these smaller truckers would increase the annual number of trips nearly three-fold: to 1347.
Sludge volumes are also dependent on weather conditions, as more silt enters waterways in heavy rain. Data from the PACl trial shows that streamwater is of exceptional clarity in the dominant dry conditions.
Managing the waste from the new system is an issue yet to be resolved by the new water authority - To Tatou Vai. Options for sludge reused or disposal are to will be subject to an Environment Impact Assessment, with an interim EIA on on-site waste disposal released in June 2021.
Water clarity is measured by a device called a turbidimeter. Values are in NTU — where a lower NTU number means clearer water.
The Sept 2020-April 2021 PACl trial data shows in the dominant dry weather conditions that Rarotonga’s streamwater is of exception clarity — under 2.0 NTU.
In April-May 2021 the project regulatory process was put under scrutiny due to attempts to fast-track the construction of two additional storage ponds in Turangi; and a temporary PACl dumpsite in Takuvaine. The proposed facilities are required as the existing sludge ponds are nearly full.
Long-term, sludge volumes can be reduced by implementing diversion systems to only collect clear streamwater.
Diversion systems check streamwater clarity (turbidity).
When collection stops, supply will be sustained by the storage tanks (2-3 days), and/or the other nine intakes.
The goal of a diversion system is to maximise water collection and optimise finished water quality; but minimise the amount of silt that enters the settling tanks.
Water clarity monitoring systems use a low-power light beam (e.g. an LED); and a sensor. In remote locations, power can be supplied by industrial battery, solar panel, or water turbine.
The light and sensor are positioned to monitor the water at a point along the treatment path. Often the monitoring is at the intake point (stream or settling tank), or post-treatment — after the sand filter.
The resulting clarity measurement is more accurate than simply clear-or-cloudy.
Controls are also computer-operated. For example, the diversion system could be programmed to stop collecting streamwater when clarity is over 10 NTU; or stop PACl dosing when clarity is under 1.0 NTU. Sustained data collection will enable the system to be fine-tuned for the specific conditions at each intake.
More sediment is washed into waterways during rain events directly after a sustained period of dry weather.
The first flush effect is more extreme in Rarotonga due to the over-steep valley topography, and the close proximity of the intake points to the central ridges. The island of Rarotonga is barely 9km in diameter, and the intakes are 3-4km inland.
Collecting only clear streamwater means less silt. Less silt means less sludge. Less sludge means less clearing and less risk of environmental degradation. Fewer trucks to the intakes. Reduced maintenance overheads, etc.
Water clarity is also improved by physical steps: screening, 3hr settling time, and sand filtration. These filtration steps would still be used when diversion is implemented. If found necessary, then additional filtration steps could be added to reliably achieve turbidity standards.
In the assessment of the Te Mato Vai system, the court expert advised that diversion is technically feasible given the existing Te Mato Vai infrastructure. The report goes on to detail the required componentry and provides an estimate for project management and construction.
When PACl is dosed, all of the Te Mato Vai sludge must be managed as hazardous waste.
The preferred method of disposal by processing sludge for agricultural re-use would require the construction of a storage/containment facility, processing plant (dewatering, and mixing the sludge with synthetic phosphorous and trace elements), and managing re-distribution.
The additional cost of sludge management has not been detailed by GHD. Agricultural re-use is deemed too costly to be viable in New Zealand and Australia. Instead PACl sludge is disposed to landfill.
If PACl is not dosed, then the sludge will be organic silt. Safe for agriculture re-use, (but still requiring storage and distribution).
With-, or with-out PACl, diversion will mean less sludge to manage.
The expert opinion was divided. No firm conclusion was drawn, but there are signs that additional investigation is warranted (and that at a minimum, turbidity monitoring systems will be required to meet the CIDWS):
“…it will be impossible to consistently meet the treated water quality requirements CIDWS without the use of a coagulant.” (p.14)
“Coagulation could only be eliminated if the water didn’t need to be coagulated during low-turbidity periods, and if during high-turbidity events if the impact of not taking water during the high-turbidity flood events that did not compromise the ability to meet demand.” (p.19)
”In order to meet the CIDWS a treated water turbidimeter is going to have to be installed anyway.” (p.20)
Referenced in Watson 2020.
In broad statements to the media, project manager GHD has tended to overlook that diversion is proposed as an additional step in the treatment process, and not as a stand-alone method of water treatment.
The claimed need for mains power has also been disputed by Watson, with a combination of battery and solar previously diployed to power a similar system in a remote location (in Northland, New Zealand).
Rarotonga’s public water supply has unique characteristics — including an unpolluted and naturally filtered source. All ten intakes feed into the same network. If one treatment facility is offline, then demand can be met by the other sites. Microclimates differ valley-to-valley.
The current sludge crisis warrants revisiting the expert opinion.
An EIA on the operational-use of PACl was announced in January 2020, with New Zealand firm Tonkin+Taylor (T+T) appointed.
An interim report was released to the public in June 2021 detailing the onsite discharge of wastes (sludge and chemically-treated water). This is to enable the new water authority to secure an operational permit. (The permit held by the Te Mato Vai Project Management Unit for construction of the infrastructure expires 30th June 2021.)
Water Treatment Plants: On-site Discharges Environmental Impact Assessment Tonkin+Taylor, May 2021
A list of applicable policies, laws, and regulations was provided as part of the Te Mato Vai Stage 2 EIA.
T+T have omitted reference to:
Both documents have direct bearing on the operation of the new system. Emphasis is instead placed on the National Sustainable Development Plan 2016-2020; and the Cook Islands National Integrated Water Resources Management Policy 2014 (IWRMP).
The effect of this bait-and-switch is to evaluate PACl-use against the overall project goal of improving the public water supply, while minimising operational considerations that impact on public health and freshwater ecology, given the lack of long-term research on the non-standard method of operation.
Material including sludge or other by products from water treatment plants is Hazardous waste.
…prohibits hazardous waste to be directly or indirectly deposited or discharged, or to seep, into a waterway.
- Public Health Act 2004
The National Water Policy refers to the same set of Policy Principles as the IWRMP, including:
Taking active measure to prevent serious or irreversible environmental damage or degradation whether the consequences are uncertain or not.
Sustainability & Environmental preservation
Strive to protect and avoid adverse interference with the natural ecology of the Cook Islands’ waters and lagoons and ensure that usage and management of water will be conducted in a manner that will preserve and enhance this resource for generations to come.
Mitigation of the impact of Infrastructure and Development on Ecological Flows
In designing and delivering future water supplies, infrastructure, services and operations will be provided in a manner that ensures that ecological flows are maintained in natural streams and waterways to the greatest extent practicable in order to ensure the protection of freshwater and coastal ecosystems.
- Cook Islands National Water Policy (2016)
Consideration of diversion systems is superficial, making no reference to the evaulation directly provided in the Watson assessment.
Disadvantages of diversion + no coagulant
- Requires sensors to measure water quality and flow.
- Requires remotely actuated valves to control flow to the treatment plant.
- These require power, which is not supplied to the plant.
- Does not address lighter suspended materials (incorporated in ‘normal flow’) including microbiological contaminants in the filter stage.
- Additional time required for removal of heavier suspended materials. This translates to larger or additional settling tank.
- Water Treatment Plants: On-site Discharges Environmental Impact Assessment. Tonkin+Taylor, May 2021
Risks to ecology are paraphrased, while countering-reassurances are provided in detail:
Aquatic flora and fauna in the streams may be adversely impacted by the discharge of water containing residual PACl from the scour ponds directly in the streams.…
Overall, the impacts on freshwater ecological values associated to the use of PACl are considered low and acceptable. (IBID)
This requires the public to cross-reference the Te Mato Vai Aquatic Ecology Baseline Report, (provided as an appendix), to appreciate specific concerns which include Reduction in diversity and abundance of macroinvertebrate communities, potentially limiting their availability as a food source (GHD Nov 2020).
Suitable control hardware, and an off-grid power source were addressed by Watson, March 2020. Noting that chlorine dosing systems also require a power source, and that industrial batteries have alreay been purchased and installed for this purpose.
Throughout the EIA, the authors conflate coagulation with disinfection. PACl is not itself a disinfecting agent. Although the settling-out of particles will reduce the levels of microbes in the water, PACl is better thought-off as enabling chemical-filtration. It helps remove, but does not kill or deactivate microorganisms.
The six month PACl trial did not conduct A/B testing to collect empirical data on the relative contributions of the PACl and physical filtration methods to improving turbidity and colour. This was despite a combined request from community group Te Vai Ora Maori, along with Avana and Turangi landowners.
The alternatives evaluated by T+T are:
Evaluation of alternative methods is hamstrung by presuming the lack of mains power. Not included is the potential benefits of PACl + diversion, especially given the issues identified by Watson with the manual / fixed dosing method, and that turbidity monitoring is a likely requirement of the CIDWS.
GHD have concluded that the chemical dosing will range from 15-35mg/L depending on water conditions (May 2021). Diversion systems can be used to limit dosing to the lower value, potentially halving the annual chemical cost.
Confirmed is Watson’s assessment that diversion is technically feasible. The Te Mato Vai system (already):
…takes water from the stream whenever it is flowing however, when storage reservoirs … are full, a series of valves will divert the water back to the stream without treatment (i.e. prior to entering the settling tank). (T+T, May 2021)
Options evaluated for managing supernatant discharge from the tanks and ponds are a) to the stream (which is noted as the preferred option); b) discharge/return to the settling tank inlet and; c) land discharge.
Dissolved aluminium entering the stream via the discharge will associate with sediment in the stream to form flocs that will settle in the region of the discharge. (IBID)
Benthic smothering is a concern raised by TVOM with greater impact in the low-flow / lo-energy stretches below the discharge point(s). Streamlife including koura burrow into the streambed to survive periods of drought/dry out.
The removal of sludge from the settling tanks (via the mid level drain) include discharge of supernatant to the stream can be scheduled to ensure any resulting discharge from the sludge ponds takes place at times of acceptable flow in the stream. (IBID)
Contrary to this advice, the removal of sludge is dependent on sludge cleaning and (automatic) sand filter backwash cycles. If there is insufficient space in the pond to contain the wastes, then the treatment process must stop. If treatment does not stop, the risks are degraded finished water quality or overflow of the ponds. In June 2021, To Tatou Vai began removing sludge from the Turangi to Papua sludge ponds for this reason.
A further supernatant management option (one not considered by T+T) is a gravity-fed cyclic wetland, to enhance the drainage potential / increase the water processing ability of the storage ponds.
…we note that while a number of people are strongly opposed to the addition of chemicals into the water treatment process,… (IBID)
Based on the T+T EIA, a reader would likely perceive there to be only minor objection to proposed method — by a handfull of citizens. And to borrow the phrasing of the PM, that such views are those of ‘extremists’. This appraisal seeks to minimise two public petitions (collectively receiving over three thousand signatures), numerous public protests, an ongoing landowner dispute, frustrated enquiry by the Ombudsman into the project permitting process, a formal grievance, and objection to the To Tatou Vai Bill. All these activities predate publication of the T+T report.
When it comes to a consideration of the social and cultural impacts, viewing the issue through the lens of the administration (comments on the Te Mato Vai Facebook Page), has severe limitations.
As with the treatment method, the Tonkin+Taylor EIA is insufficiently localised.
After the release to landowners of preliminary reports on the Sept 2020-May 2021 PACl trial, Te Vai Ora Maori requested additional data analysis.
|Turdibity||No. of samples|
|Under 1 NTU||868|
|Between 1 and 2 NTU||507|
|Between 2 and under 5 NTU||263|
|5 or over NTU||125|
Source: GHD June 2021 (pers. comm.)
Of the 1763 raw streamwater samples collected at nine of the ten intakes, Sept 2020-May 2021:
Note that disinfection is a separate treatment step, coming after settling and/or coagulation. Improving water clarity prior to disinfection improves the effectiveness of UV or chemical disinfection methods.
If collection were only to stop when streamwater turbidity is over 5.0 NTU, then 93% of the available streamwater would be collected for treatment.
Overnight, when water demand drops and when the storage tanks are full, treatment already stops and water is diverted back to the stream.
The coagulation process, settling time, and sand filters further improve clarity. This means that the diversion turbidity threshold could likely be raised to increase abstraction rates — without compromising finished water quality.
The PACl trial data seems to largely answer a concern raised in the expert assessment:
Coagulation could only be eliminated if the water didn’t need to be coagulated during low-turbidity periods, and if during high-turbidity events if the impact of not taking water during the high-turbidity flood events did not compromise the ability to meet demand.
— Watson, March 2020
Watson was then futher engaged to cross-examine the PACl trial report; and provided a response to landowners in September 2021. Noteably this expert advice was received after To Tatou Vai had submitted application(s) to the Rarotonga Environment Authority seeking permission to discharge sludge to onsite ponds and residual to the streams adjacent to the water treatment facilities; and for ‘temporary’ sludge disposal sites. It is also unlikely that the members of the permitting authority were provided with the expert report as part of their deliberations.
In the desktop review Watson raised a number of concerns regarding the results of the trial, most critical being the failure of the treatment system to reliably achieve the required turbidity standard.
The trial has shown that gross excursions of treated water turbidity can occur under flood conditions. These are not consistent with the project’s objectives. Various mitigation measures need to be considered (including raw water diversion) and the best one(s) implemented.
— Watson, Sept 2021
An enviroinment impact assessment (EIA) report was submitted to the Rarotonga Environment Authority in May 2021. Authored by New Zealand consulting firm Tonkin+Taylor, this report seeks permission for the ongoing operational discharge of PACl sludge to the onsite storage ponds; and chemically-treated water to the streams adjacent to the ten water treatment facilities.
Te Vai Ora Maori provided a response to the EIA, raising additional environmental considerations that were not acknowledged by the GHD ecology report; and proposing that operational processes and pond design could be altered to avoid the need for stream discharge.
August 2021, in reply the water authority refuted the merits of implementing diversion with the following broad points:
Foremost, water treatment (should be) designed based on source water quality. In practical terms, this starts with the investigation of catchment areas and assessment of the raw water quality. In engineering terms, applicable standards are then identified for the treatment system to achieve. The engineers responsible for specifying Te Mato Vai system have failed to match the treatment method to the context of a remote and unpolluted tropical island — let alone consider ecological and cultural values.
Referring back to Watson’s review: without modification, the Te Mato Vai system is unable to consistently achieve the required turbidity standard. Watson was also critical of the casual understanding exhibited (by GHD) regarding water standards and quality assurance:
I am concerned that the [PACl trial] report is suggesting that the results of the trials “…are considered when setting the upper treated water turbidity compliance value” in the finalised CIDWS. Accepted practice is for water treatment plants to be designed and operated to meet a standard. …
The report does not address, or even consider, the gap between the original treatment goals and the results of the trial. While I could accept the occasional excursion up to 5 NTU (including in the post-backwash periods) there are several data points from the trial that indicate the loss of control of the treatment process. There is no attempt to understand these better and propose mitigation measures.
— Watson, Sept 2021
The water authority’s response seems largely an attempt to shore-up the established defense against improving the system — without any genuine consideration of the merits of modification. Aside from the minor concessions made to landowners in terms of intake site layout, this form of reponse is consistent with the authoritarian stance adopted by administrators over the life of the project. The mantra is to reject, refute, discredit, or ignore any criticism. Long-term infrastructure decision have been recast as a team sport, where ‘wining’ is more important that meeting the project goals.
The offshore consultants engaged by the Cook Islands Government have reduced consideration of diversion to a binary decision. Implementing diversion does not mean discarding any of the other treatment steps. Nor is diversion the only method of improving treatment efficacy: catchment rehabilitation; construction of a weir upstream of each intake; and pre-filtration of streamwater would also improve turbidity levels.
Classification/subjects: Te Mato Vai, To Tatou Vai, coagulation, flocculation, water treatment residuals, sludge, polyaluminum chloride, PAC, PACL, diversion, water standards, water quality, Rarotonga, Cook Islands, South Pacific, environmental impact assessment, EIA.
Working Draft. Updated: 8 Dec 2021.