Sat, 4 Jun 2016 at 4:11 p.m.
● 30 Nov 2015 (McElhone)
Infrastructure Performance During June 2015 Flood Event | Meeting Video
● 20 Apr 2016 (Media Release)
Report on South Dunedin infrastructure performance during June 2015 flood released
● 26 Apr 2016 Agenda (and reports) Infrastructure Services Committee
Item 5 (Stokes, pp 6-27) South Dunedin Public Infrastructure Performance during June 2015 Flood Event Follow up | Meeting Video
The following content from consulting engineer Neil Johnstone is provided for your information and convenience. However, the site owner cannot accept any liability for its accuracy or content. Visitors who rely on this information do so at their own risk.
Received: 18 May 2016 [full text]
An Independent Review of DCC Report : ‘South Dunedin Public Infrastructure Performance during June 2015 Flood Event Follow up’
By N. P. Johnstone, MIPENZ
1. This review complements my peer review of DCC’s first flood report, published in November 2015. This review assesses the content of the second report (described henceforth as “the report”) published in late April 2016, and contextual statements made elsewhere by DCC staff and elected members. The author of the report is Ms R. Stokes. The technical qualifications and relevant experience of the report’s author are not stated.
2. I consider that there is a need for such a review for reasons of historical accuracy and context, the identification of solutions (which can only be achieved be if the problem is understood and acknowledged) and – most importantly – to provide a considered assessment of what South Dunedin’s current flood risk really is, noting that two events (of which only the recent one caused major inundation) in five decades does not suggest a current flood risk much different from that existing in many other established New Zealand communities, despite some landuse changes. It is emphasised that the flood had nothing to do with climate change, nor therefore does this review. The failure to understand the issues may lead to inaction or to inappropriate and expensive actions.
3. This review may be criticised for being repetitive on some issues, but the repetition is at least partly driven by the number of times challengeable information on the flood event and its causes has been circulated by DCC. In many respects, the report under review could be seen as a concentration of such challengeable information. The report is solely based on my research, knowledge and experience; any errors are therefore mine, but hopefully, few.
4. This review has led to the following conclusions:
4.1 Council’s continued insistence that the June 2015 rainfall event was the largest since 1923 remains erroneous;
4.2 Pre-existing groundwater levels were unexceptional, and had no impact on the flooding, contrary to claims made in the report, previously and subsequently;
4.3 South Dunedin does not have a significant imminent exposure to stormwater flooding. This finding is based on the original design parameters, historical performance, an absence of groundwater issues, and provided existing infrastructure is properly maintained, monitored and operated;
4.4 Problems at the Portobello Road Pumping Station caused elevated flood levels and prolonged the period of inundation, but the report acknowledges only the latter;
4.5 Similarly, the now-admitted failures to ensure that mudtanks were properly maintained impacted adversely on flood levels attained in some locations at least, and prolonged the period of inundation in many areas;
4.6 Comparisons with the 1968 flood event can be instructive in assessing the impact of Council failures in 2015 in terms of water level, disruption and cost. The report fails to make such assessments.
5. My review of the first report, written by Ms L. McElhone, was driven by DCC claims that the prime causes of the flood were high sea and ground water levels, a 150-year, then a 100-year, then a 63-year (and incidentally and extraordinarily now again a 100-year*) rainfall event, and confirmed that Portobello Road pumping station issues added not less than 200mm to peak flood levels. That review also demonstrated that the rainfall event of March 1968 was demonstrably larger than that of June 2015, but caused much less damage, and that land use changes added up to 150mm to flood levels (based on DCC’s unconfirmed data on impermeable areas). Any consented landuse changes should, in my opinion, have been compensated for in past years with additional infrastructure to maintain drainage standards and South Dunedin’s protection standards.
(*Ms Stokes to John Campbell on Checkpoint, 21 April 2016).
6. Exaggerated assessments of both the historical significance of the 2015 rainfalls and groundwater levels, and the absence of mudtank information originally helped DCC promote its position of zero liability. The mudtank maintenance failures are at last largely revealed in the new report; significant mudtank maintenance issues were previously reported by Cr Lee Vandervis as early as 2014, but were seemingly largely ignored by DCC. Paragraph 37 of the report which reads: “Mudtank maintenance and performance in general has been the subject of focus for a number of years”, appears vague, and therefore requires elaboration. The statement, if accurate makes the failures more disturbing. There is still some unfortunate reliance on the groundwater myth (paragraphs 2 & 32), and to the underestimation of the 1968 event (paragraph 20). There appear to be newly-entrenched positions at DCC that the existing stormwater system is inadequate, presumably based on the report’s paragraphs 23-27, and reinforced in recent public statements from Ms Stokes and Mayor Cull, that the flood would have occurred (or that a serious flood would have occurred) even if the current system including the mudtanks had operated at optimum. This review strongly disputes such claims, and uses the well-documented event of March 1968 as a very useful “model” for key comparisons between a contained flood (1968) and a disaster (2015).
7. The general understanding was that DCC’s second review was to concentrate in detail on the performance of mudtanks, and was to be peer reviewed. Detailed reporting and peer review processes were understood to be the reason for the extraordinary delays in publication. In reality, only paragraphs 33-61 deal with mudtanks issues and no peer review is included. It is noted that Mayor Cull confirmed to John Campbell on Radio NZ’s Checkpoint programme (22 April 2016) that peer reviews of the report had been produced. The peer review(s) could usefully have been attached to the report; failing that, the report’s author should have explained their absence.
8. The pump stations at Portobello Road and Tainui (Musselburgh) are noted in the report’s paragraph 12. DCC’s failures at Portobello Road are now common knowledge, although the relative contributions to that failure of screen blockage and pump switching failure remain undiscussed. It is noted that Council has made no comment on the performance of the Tainui pump in either of its reports. This is a remarkable omission for an infrastructure review and follow-up. DCC staff could provide little information on the Tainui Station’s performance in response to a question from a Councillor at the Infrastructure [Services] Committee’s meeting on 26 April. Information on the performance of this station needs to be made available.
9. Paragraph 13 of the report refers to the stated low standard of existing stormwater service across South Dunedin as being of the order of a 1 in 2- year rainfall event, due to “bottlenecks in a number of locations”. Neither the bottlenecks, nor their locations are described in any detail. These bottlenecks do not appear to have caused significant or widespread issues on anything like a 2-year basis in recent decades. It would have been helpful to know if they are substantially different now compared with what they were in 1968. In any case, any problems caused by any such “bottlenecks” should surely have been addressed as a priority long before the 2015 disaster?
10. DCC should have critically appraised in practical terms the current existing standard of South Dunedin’s drainage infrastructure in light of the known adequate performance of the system since it was upgraded around 1960. To my knowledge, there have been 2 significant flood events in the intervening five or so decades. These occurred in March 1968 and in June 2015, the latter being seriously aggravated by DCC failures. Possibly there have been other minor incidents, small enough to have been largely forgotten. This all suggests a far better actual existing standard than a 2 or 3-year ARI; 50-plus years of reality suggest an actual standard in the order of a 10-year or even a 20-year ARI or better. Flooding of South Dunedin is clearly far from inevitable on a regular basis, despite frequent statements to the contrary from DCC in recent weeks (both Mr Cull and Ms Stokes have emphasised this challengeable claim in nationwide radio interviews, as has Mr Cull on Dunedin television on 6 May and earlier to the South Dunedin Action Group on 3 May). In DCC’s May public issue of FYI, Mr Cull attributes his understanding of Dunedin’s stormwater to the report. It is noted that several of Mr Cull’s comments therein are inaccurate, and some of the inaccuracies are attributable to Ms Stokes’ report. Mr Cull’s comments make no mention of pumping station or mudtank failures.
11. By contrast, DCC’s 1968 flood report is helpful. The original stormwater infrastructure design described in that report anticipates some backing up in the streets (what we might now call inconvenience flooding) to occur on a 10-year basis, and resulting from a rainfall intensity of 19mm per hour. This is broadly the situation that occurred in March 1968, when peak rainfall intensities were recorded as 23mm per hour. In June 2015 the peak rainfall intensities did not exceed 16mm per hour, yet the flooding was historic and disastrous for many. My necessarily limited research reveals that approximately 100 houses and businesses experienced some internal flooding in 1968 whereas, according to local MP Clare Curran, approximately 1,000 houses were flooded in 2015, but to greater depths, for much longer and (in some cases) accompanied by sewage. In my view, this highlights the extent to which the 2015 flooding was substantially aggravated by DCC’s mismanagement. This mismanagement needs to be fully acknowledged, to enable a proper understanding of the issues within the community and beyond, and to ensure appropriate responses.
12. DCC has previously used Otago Regional Council determinations of rainfall return periods that are based on information that is incomplete and not well considered. A 24-hour rainfall duration is longer than should ideally be used for analysing rainfall probabilities for such a small catchment as South Dunedin, but that is the approach taken by our Councils. It is an approach of convenience, probably because tabulations of 24-hour rainfall are the most easily accessible. A 4 or 6-hour design storm is probably more appropriate for such a small catchment, as it will encompass both high rainfall intensities and significant volumes of runoff. For statistical/planning purposes, both Councils seem intent on treating the March 8-9 1968 rainfall as if it didn’t happen. The fact is that 158mm of rain fell in 1968 within 24 hours, compared with 142 mm in June 2015. Both DCC reports reproduce an ORC tabulation stating that the June 2015 was the 2nd highest recorded (after 1923). This is demonstrably incorrect. Consideration of shorter duration storms would almost certainly further lower the “ranking” of the June 2015 rainfall. The 2015 DCC report (paragraph 59) reads: “The event took place over two days on 8/9 March 1968. The official NIWA data records this event as having lower 24 hour rainfall than the 2015 event because only daily data are part of the NIWA dataset for 1968. Therefore the NIWA records show two consecutive days with moderately high rainfall rather than a single 24 hour period with very high rainfall. A Council report (1968) prepared after the event provides hourly rainfall records which give a 24 hour total at Musselburgh of 158.2mm.” DCC and ORC appear not to have appreciated the distinction between a 1-day rainfall and a 24-hour event. It is undeniable that the March 1968 24-hour rainfall was 158mm, ie more than 10% greater than that of June 2015. [Additionally and of some significance, there is a reported 24-hour rainfall of 5.4 inches (137mm) at Musselburgh in February 1892, only about 3% below the 2015 value.] This and other detailed analyses of the 1929 Dunedin rainfall event makes the ORC/DCC assessment of the 2015 event as a 63 year event unjustifiable, as my review of DCC’s first report has previously demonstrated.
13. It is particularly important that that the similarities of the 1968 and 2015 rainfalls are acknowledged. This then allows a comparison of the respective flood effects to understand and assess the consequences of DCC’s failures in the latter event.
14. Significantly, the 6-hour rainfall recorded in 2015 totalled 54 mm, whereas in the 1968 event there were effectively two separate 6-hour events totalling 65mm and 61mm respectively and with peaks about 7 hours apart. The “second” storm fell on an already wet catchment.
15. There are further concerns. Paragraph 21 of the report states: “The Musselburgh Rain gauge provides the longest continuous record within the Dunedin area and is therefore considered the most accurate for the rainfall event estimate of likelihood.”
16. This appears to contradict Ms Bidrose’s response of 7 March to my first peer review when she wrote: “Although the DCC report from the 1968 flood includes hourly data for the event, there is no ability to confirm the accuracy of the data, the total rainfall falls across two different dates meaning that it does not show as an exceptional event. It would be inappropriate to selectively modify analysis of the National dataset to replace data from a different source, hence the anomaly was noted in the Council’s November report. In summary, we do not believe that your “20-30” year return period estimate is valid.” In my view the only anomaly/selectivity is DCC’s approach to effectively purge the totally reliable 1968 rainfall data from any analysis. From a statistical approach, this is tantamount to decreeing that the 1968 event did not occur. It did, and it was indisputably a greater rainfall event than occurred in June 2015.
17. Then at the DCC’s Infrastructure [Services] Committee meeting of 26 April 2016, Ms Stokes responded to a direct question from an elected member to the effect that the June 2015 rainfall was the largest to occur in South Dunedin since 1923. This was demonstrably incorrect. Ms Stokes repeated this assertion on Radio NZ’s Checkpoint programme on 21 April 2016. She then implied that the rainfall was a 100-year event, and that no New Zealand city would have been able to cope with it. Supporting statements are not evident in her report, and ironically South Dunedin itself coped adequately with a larger event in 1968.
18. In paragraph 27 of the DCC report it is emphasised that runoff exceeded the Portobello Road pump capacity for a total of 17 hours during the June 2015 flood. This was re-emphasised by Ms Stokes in her interview with Mr Campbell. I have calculated that this total “excess rainfall” across those 17 hours was of the order of 50mm. I have taken a similar approach to analysing the March 1968 flood. Similarly calculated “excess rainfall” fell for a lesser period (13 hours), but the total “excess rainfall” was of the order of 90mm due to the far higher rainfall intensities experienced then (up to 23mm per hour). Even from this perspective, the 1968 event was clearly the larger. In my view, this comparison should have been presented in the report for balance.
19. It is noted that the claim of 17 hours of “excess rainfall” (ie runoff in excess of what the pumps can handle) originated in Ms McElhone’s report (her paragraphs 36 & 37). Ms McElhone’s conclusion that there was such a prolonged period of excess rainfall is entirely flawed in my opinion as her calculations are based on the unjustifiable premise that there was “little or no” infiltration throughout the rainfall event. It is known that there was substantial infiltration because there were significant rises in groundwater levels at all of ORC’s bore levels once the rainfall commenced. Nor does the fact that groundwater levels reached ground level at some of those bore locations necessarily indicate a cessation of infiltration; throughflow would have continued, though possibly slowed. A more accurate assessment of the period of excess rainfall is demonstrated in paragraph 30 of Ms McElhone’s report. Here, the original design assumption of normal infiltration as described by Ms McElhone means that 9mm of rainfall could be accommodated by the pumps, because infiltration took care of about half the rainfall. Allowing for realistic infiltration reduces the true period of excess rainfall in the 2015 event to just a couple of hours around midday on 3 June. That is the real rainfall excess, and that is the only rainfall that should have been expected to cause more than minor flooding in June 2015 if the system had been maintained and operated properly.
20. DCC has highlighted the fact that there were two peaks in rainfall in the 1968 event, and that the system therefore had time to partially recover prior to the second peak. The lull (not cessation) in rainfall was only 3 hours long, and any dropping of water levels was only possible because the infrastructure was working efficiently, and (presumably) infiltration was occurring. This needs to be compared with the inability of surface water to drain quickly (or at all) after rainfall eased after midnight on 3 June 2015. We now know a very large proportion of mudtanks were blocked, and that there were ongoing problems at the Portobello Pumping Station.
21. It is stated in paragraph 31 of the report that the poor performance of the Portobello Road pumping station contributed to the length of time of flooding, but the depth of flooding is not mentioned. The depth of flooding determined the extent and manner of damage and disruption. This increase in depth of flooding due to Portobello Road failures has previously been assessed both by myself and DCC to be of the order of 200mm. Excluding reference to the depth of flooding created by pumping station issues is unfortunate. The report therefore extraordinarily fails to acknowledge any impact on flooding depth to either mudtank or pumping station issues.
22. Ironically, in paragraph 32 of the report it is stated that ground water levels were “elevated” prior to the rainfall event, but again the report fails to mention that the degree of prior groundwater “elevation” was very minor compared to the rises experienced once the rain started. Real, significant increases in flood depth costing many millions of dollars and disruption are not acknowledged, but inconsequential ground water levels are. This rise in ground water level during the rainfall event is indicative of steady-not impeded-infiltration. The paragraph concludes with the statement that “little, if any of the rain was able to be absorbed by the ground for some time during and immediately post the flood”. The fact that ground water levels rose steadily from the onset of the rain in the manner recorded by ORC demonstrates the opposite, ie that there was no abnormal impediment to infiltration resulting from pre-existing ground water levels. My summary of actual ground water levels recorded is reproduced from my review of DCC’s first report here:
23. [Groundwater level data relied on by the report is sourced solely from ORC bore installations at four South Dunedin locations (Kennedy Street, Culling Park, Tonga Park and Bathgate Park). The data is produced graphically and descriptively in the ORC report (pages 25-30 and figures 21-24). The DCC report states: “During the June event, observations from the Otago Regional Council of the groundwater conditions prior to and during the event show that the groundwater was elevated prior to the event and actually reached the surface at several of the South Dunedin monitoring bores.” In my view the statement is inadequate as it fails to clearly represent the magnitudes, timing and cause of groundwater level rise. The ORC report in fact indicates that:
● GW levels at Kennedy Street were just 71 (674 minus 603)mm above average before the flood event, but rose 997mm during the event, and peaked 261mm above ground level.
● GW levels at Tonga Park were just 87 (644 minus 557)mm above average before the flood event, but rose 722mm during the event, and peaked 346mm above ground level.
● GW levels at Culling Park were just 123 (198 minus 75)mm above average before the flood event, but rose 707mm during the event, and peaked just above ground level (by 77mm).
● GW levels at Bathgate Park were just 62 (751 minus 689)mm above average before the flood event, but rose 238mm during the event, and peaked a full 532mm below ground level.
The small rises in groundwater levels before the June 2015 rainfall event are in all four cases attributed by ORC to minor rainfalls that occurred on 12 and 26 May. The only sensible conclusion is that ground water levels are determined by rainfall and relatively rapid infiltration.]
24. Paragraphs 33-61 deal with the issue of mudtank performance. The belated admissions that mudtank maintenance was inadequate came as no surprise to residents and business operators of South Dunedin, but it appears to have come as a surprise to DCC. This review focuses on paragraphs 54 and 55 of the report which attempt to minimise the consequences of the mudtank blockages. The statement in paragraph 54 to the effect that there may (my emphasis) have been an adverse impact on the time for water to clear as a result of mudtank blockages seems a grudging partial acceptance of a greater issue. Of course clearance was delayed. Paragraph 55 which claims: “However, given the volume of rainfall and the system at capacity, the water would have been unable to enter the network even if all mudtanks were clear”, is strongly challenged. In a properly operating system there would certainly have been a period when the drainage network was full, and some surface flooding would have occurred, generally in the manner of the March 1968 event. Regrettably, paragraph 55 has seemingly been relied on by DCC representatives to issue statements to the effect that the flood (rather than some flooding) would have occurred regardless of the state of the mudtanks. Such conclusions are disputed for the following reasons:
25. No consideration appears to have been given in the report to the extent of pumping station screen blockage that was caused by material passing from overfull mudtanks into the stormwater network because the mudtanks’ trapping ability was compromised. This would presumably have contributed to the poor performance of the pumping station, virtually from the commencement of the event.
26. The contribution from adjacent catchments such as St Clair would have, in my view at least partly consisted of water by-passing blocked or partially blocked mudtanks there, before adding to South Dunedin’s problems.
27. The mudtank blockage certainly delayed the flood’s recession, but the issue is not solely one of time. For every minute that ponded water could not drain through blocked mudtanks adjacent to roads etc, nearby properties were susceptible to ongoing spillover via gravity and aggravated by largely uncontrolled traffic movement. All of this would have ensured there was continued building of flood levels across private and business properties, inside and out.
28. I have concerns with the assertions made in paragraph 27 of the report that rainfall effectively exceeded the capacity of the pumping station (4.1mm/hr of rainfall) for a period of 17 hours. Intuitively, this cannot be correct as flooding of South Dunedin would be occurring on a frequent basis if it were. The 17 hours of excess rainfall claim originated in DCC’s first Infrastructure report, authored by Ms McElhone. Her analysis relies primarily on the unsustainable position that there was no infiltration of rainfall. ORC’s bore water level data demonstrates a different reality.
29. The above considerations have led me to the conclusion that the report is technically challengeable, inadequately researched and lacking the assurance of independent technical peer review, all to the ongoing detriment of the victims of the 2015 disaster, ie the residents of South Dunedin. Reliance on the findings of the report cannot therefore be recommended.
30. I believe an independent audit of DCC performance prior to, during and in the lengthy period following the flood event would prove beneficial in achieving the right outcomes for South Dunedin.
Related Post and Comments:
9.5.16 South Dunedin: Fixing Council attitudes and badly maintained…
█ 6.5.16 South Dunedin Action Group: Notes of meeting with DCC (3 May 2016)
29.4.16 Vandervis emails batch 2 #Dunedin #infrastructure #flood #mudtanks
27.4.16 Vandervis emails batch 1 #Dunedin #infrastructure #flood #mudtanks
█ 27.4.16 DCC meeting and apology NOT Enough— #SouthDunedinFlood
21.4.16 Planning for South Dunedin, but wait….
█ 20.4.16 DCC Politics : Release of Infrastructure Report #SouthDunedinFlood
14.4.16 South Dunedin flood risk boosters #ClimateChangeCrap #PissOffPCE
1.4.16 “O Oysters, come and walk with us!” #SouthDunedinFlood
█ 31.3.16 DCC: Infrastructure report 2 pending —Mudtanks & stormwater…
16.3.16 Cr Andrew Whiley easily points up Greenie Hawkins’ inadequacy…
█ 8.3.16 Johnstone independent review of DCC report…
2.3.16 DCC compels extensions on LGOIMA requests #SouthDunedinFlood
26.2.16 Mudtanks and drains + Notice of Public Meeting #SouthDunedinFlood
21.2.16 DCC…report (30.11.15) subject to ‘internal review’ only…
13.2.16 South Dunedin Flood (3 June 2015): Bruce Hendry via ODT
4.2.16 2GP commissioner appears to tell Council outcome… #hazardzones
4.2.16 Level responses to Dunedin mayor’s hippo soup #Jun2015flood
30.1.16 DCC Rates: LOCAL CONTEXT not Stats —Delta and Hippopotamuses
25.1.16 DCC: South Dunedin Integrated Catchment Management Plan (ICMP)
19.1.16 Listener 23.1.16 (letter): South Dunedin #Jun2015flood
16.1.16 NZ Listener 16.1.16 (letter): South Dunedin #Jun2015flood
10.1.16 Infrastructure ‘open to facile misinterpretation’…. or local ignore
5.1.16 Hammered from all sides #fixit [dunedinflood Jun2015]
24.12.15 Site notice: posts removed
3.11.15 South Dunedin Flood | Correspondence & Debriefing Notes … #LGOIMA
█ For more, enter the terms *flood*, *hazard* or *south dunedin* in the search box at right.
Posted by Elizabeth Kerr
Election Year —this post is offered in the public interest.