Closing submission to Public Enquiry - Tim O’Hara re Costs

Closing submission to the CPO inquiry for the Oxford Flood Alleviation Scheme

Tim O’Hara

26 January 2024

INTRODUCTION

1. Thank you, Madam Inspector, for allowing me the opportunity to make these closing remarks.
2. In my written submission to the inquiry, I argued that the some of the CPOs should not be confirmed; these being in respect of the land sought for the flood channel between Seacourt Nature Park and Old Abingdon Road.
3. After having seen or heard all the evidence, I am even more strongly of this opinion, and I’m going to explain why.

BACKGROUND TO THE CPOS

4. Firstly though, a word on the CPOs. They were made to create the main part of a 5km long flood channel, which is just part of the proposed flood alleviation scheme.
5. Other parts include: raised flood defences at four key locations; and measures to allow flood water to escape from the floodplain by going beneath the railway and three roads. The EA would also provide new bridges to cross the new channel at six locations. I will come back to this.
6. The scheme as a whole is intended to provide long-term flood alleviation, largely prompted by a series of floods between 2000 and 2013/14 (EA/1a 2.3).
7. However it’s important to understand that, since mid-2014, many flood alleviation measures have already been put in place (EA/1a para 5.9(a)). There is ample evidence that these, together with others works in 2015 and 2016 (OX023N/1a page 2), have significantly improved flood protection in Oxford. So, it’s against this existing level of protection that any new scheme should be gauged.

REQUIREMENTS FOR CPOS

8. In my evidence (OX002N/1a) I referred to the government publication “Guidance on compulsory purchase process and Crichel Down rules” (2019 edition). This provides the framework for making CPOs, and considering appeals against the same.
9. Taken together, section 12 and 12 of the guidance mean that

• Confirmation of a CPO requires the minister (or her/his representative) to test whether there is a compelling case in the public interest; and that
• The wider public interest is distinct from the intention of the acquiring authority, and the concerns of those with interests in the CPO lands.

In this instance it provides the justification for looking at costs and benefits that go to the wider public interest, and not just those the EA have considered.

THE TWO KEY ISSUES

10. I believe there are two key issues related to the EA’s proposed flood alleviation scheme.
11. First, there is strong evidence that the main part of the channel should be omitted from the scheme, based on the balance of costs and benefits. I will signpost this evidence, and explain why this removes the need for the channel, and therefore the CPOs.
12. Second, the EA have argued that the scheme must include the full planned length of the channel, and have provided reasons that go beyond the very limited flood protection benefits shown in their own modelling. So later I will outline the EA’s reasons, and show why they are flawed, unconvincing and inadequate.

BENEFIT COST ANALYSIS OF THE PROPOSED SCHEME AND ALTERNATIVE A

13. Starting with benefit and costs… In arriving at their choice of scheme the EA have identified two types of costs and benefits: those that can be included within the formal benefit cost analysis (BCA), and those that cannot.

14. Regarding the former: the BCA provides present day money values of 100 years of flood protection benefits, and the costs of providing these.

15. Benefits are defined as the flood damages that are avoided by the option being assessed. Therefore this measure takes account of both the severity of damage expected in a given flood event, and the probability of such an event actually taking place.

16.Costs within the BCA are limited to the direct cost of the option being assessed, including construction, maintenance, and land acquisition.

17.The EA have modelled the proposed scheme. They have also modelled how omitting parts of the channel would alter the results of the BCA. Firstly Alternative A1, which omits the channel between Seacourt Nature Park and Willow Walk (i.e. in Hinksey Meadow). And secondly Alternative A2, which omits the channel between Seacourt Nature Park and Old Abingdon Road. The results of this modelling (CD3.29, Table 5, page 16) show that the total benefits, benefit-to-cost ratios and net present values of Alternatives A1 and A2 are comparable to the proposed scheme.

18.For example, the proposed scheme has flood protection benefits only 2.4% higher than A2. Moreover the higher cost of the channel means that the net present value (NPV, total benefits minus total costs) of the scheme is only 0.8% (£11.3 million) higher than A2 as it is currently modelled (CD3.29 Table 5, page 16).

19. However, the modelling of A2 and A1 was on the basis that no substantive changes were made to the scheme other than omitting the channel (CD3.29, page 2). This means that the other flood alleviation measures within A1 and A2 were not optimised to reflect the omission of the channel.

20. Regarding optimisation: there is evidence from Mr. Coombes regarding A1 (summarised in INQ/40), and Mr. Carpenter regarding A2 (OX027S4A) that the flood protection benefits could be increased by simply increasing the height of some of the flood defences.
21. There are other examples of optimisation of the alternatives that the EA could have carried out but haven’t. The proposed scheme includes the construction of several bridges which are needed in order to span the new channel. The modelling of bridges is included within A2 even though four of them are in locations where A2 omits the channel.
22. These bridges are Willow Walk, North Hinksey Causeway, an unnamed footbridge to the south, and a bridge at Devils Backbone (INQ/12 and CD2.15). I do not know what they cost, but they will cost something, and omitting them would reduce the cost of this option, which in turn would increase its NPV and benefit-cost ratio.
23. There is another way that optimising A2, and this could improve the flood protection it offers.
24. In my cross examination of Mr Harding (30/11/23) he confirmed that Willow Walk acts as a barrier to flood water moving from north to south. If the water cannot escape under Willow Walk, there is a risk to properties either side of Botley Road. For this reason the EA installed six culverts under Willow Walk in 2011.
25. The way the EA has modelled Alternative A2 means that these culverts would be removed and replaced by a new bridge. Following my cross-examination of Mr. Harding (30/11/23), he has provided a drawing and additional information about this bridge (INQ/42).
26. The drawing shows an excavation for the channel where it passes under the bridge. However, in the way the EA modelled A2, the channel and the corresponding area of excavation would not exist. Water would only flow under the bridge when it was at ground level in Hinksey Meadow which is 55.7m AOD^[1].
27. At present, flood water can drain from Hinksey Meadow through the culverts at 55.0m AOD, so 70cm lower than the ground level at Hinksey Meadow, via a ditch between Willow Walk and Hinksey Meadow.
28. The EA’s modelling of A2 means that more flood water than at present would build up in Hinksey Meadow before it is able to get under Willow Walk. Clearly this is at odds with the efficient conveyance of flood water.
29. 29. So optimising A2 at Willow Walk– which the EA haven’t done - might involve adding more culverts set at the same level as the existing ones. This would surely be cheaper than building a large new bridge AND provide better flood protection than what the EA has modelled.
30. If optimisations led to the cost of A2 being reduced or its benefits improved, by just £11.4 million, then this would raise the net present value of A2 above that of the proposed scheme. This in turn would effectively make it the preferred choice within the BCA.
31. However, even if this wasn’t the case there are other costs associated with the channel that should be taken account of when considering the wider public interest.

COSTS ATTRIBUTABLE TO THE CHANNEL ALONE

32. Firstly, there is the certain loss of 1.33 of nationally rare, irreplaceable MG4a grassland in Hinksey Meadow, and risk to the remaining parts of the meadow. I believe Mr Byass and Mr Lyness will refer to these in more detail. Accompanying this is what Dr Rod Chalk estimates as the loss of 3780 mature trees (OX027S/1a, page 21), with related impacts on biodiversity, insect and animal habitat.
33. Second, the loss of amenity caused by the construction of the channel.
34. The EA have undertaken to maintain access to rights of way, Hinksey Meadow and other open space. However the work areas would have to be out of bounds for safety reasons, and this would reduce the space that could be accessed. Moreover, exchanging an intact iconic green space with open views of a giant construction site must comprise a loss.
35. Third there are the impacts on the A34. The construction of the channel would require excavation of 450,000m3 of material, most of which is for the section of channel between Willow Walk and Old Abingdon Road. Much of the spoil for this would have to be removed by means of 25,000 lorry loads onto the A34 (and again 25,000 off the A34) via two slip roads at South Hinksey. Therivel’s evidence (OX002N/1a page 15) sets out the issues associated with this.
36. One is the safety risk presented by 25,000 heavily laden lorries accessing a busy trunk road at less than 10mph from a standing start (OX002N/1a, para 45-62). These impacts cannot be readily quantified in money terms, but they pose a safety risk that would affect anyone using this road during the three year construction period.
37. Another issue is the economic cost of traffic delays that stem from a 40mph speed limit for the three years of construction (OX002N/1a, Figure 3.10). These costs can be quantified and monetised (para 63-71). Dr. Therivel uses Department of Transport methodology to calculate the cost to road users based on delayed journeys.
38. At para. 71 she concludes:

“The net benefits of the OFAS channel is expected to be £11.3 million. The cost of only slowing down traffic during the excavation of the channel would be £10.56 - £34.96 million, i.e. between 93% and 309% of the benefits that the channel is expected to bring over 100 years…”

45. Although the EA dispute these findings, they have provided no estimate or calculations of road delay or its impact. It’s as though they are indifferent to these costs. Indeed, when Dr. Therivel cross examined Mr. Lear for the EA (30/11/23), he suggested that it would be possible to deal with any unexpected traffic issues by reducing the speed still further, to 20-30mph. Obviously this would further increase the cost of delays to motorists.
46. The section of the A34 affected by the scheme is not only part of a nationally-important trunk road. It is also part of the city ring road, and therefore it is important for it to flow freely for the functioning of the city itself. This is especially so given the city’s endemic traffic problems.
47. Hence it is galling that the EA’s evidence (e.g. EA/3a para 9.10, and EA/3c.1 page 186) talks about the value of keeping Oxford open for business, and makes claims for how the scheme will do this by keeping open transport links. Yet on the other hand the cost to the city of three years of reduced speed on the A34 merits no assessment or even a mention.
48. The fourth cost of the channel stems from the opportunity cost of the extra money needed to construct it.
49. The EA’s modelling of A2 (CD3.29 Table 5, page 16) shows it costing £24 million less than their proposed scheme. Moreover, as I pointed out earlier, there are grounds for believing an optimised A2 would save even more money.
50. During this inquiry a National Audit office report (INQ/17) came to light, which states that the value of the existing £5.2 billion fund for flood and coastal defence has been eroded due to inflation. Previously it had been expected to provide protection for 336,000 dwellings, but price inflation has reduced this to 200,000.
51. In my cross examination (30/11/23) of Emma Formoy, I asked if money not spent on this scheme would revert to the central fund and she agreed. Hence money saved on the Oxford scheme would make it more likely that another scheme elsewhere could be funded. Surely saving £24 million to help other schemes is something that you might expect the EA to support.

OTHER BENEFITS OF THE SCHEME

46. Whilst I’ve referred to the costs of the channel that are not included within BCA; the EA have also sought to justify their scheme by identifying benefits that are not included within the BCA. I refer to the main ones below.
47. The EA (EA3/a, para 9.10) refer to local economic benefits over 100 years of the proposed scheme compared to the measures in place in 2016. These benefits are excluded from the BCA because they relate to local effects of economic loss which might be displaced as economic gain outside of Oxford. The EA estimate the benefits from economic losses due to flooding of business property to be £102 million, and those from avoidance of transport disruption as £52 million.
48. However the EA has provided no evidence to show that Alternative A2, let alone an optimised A2, would not also prevent these costs. Indeed, the flood maps that compare the proposed scheme with A2 (CD3.29 Figures 3-6, pages 10-13) show very little difference to flooding of commercial areas and road/rail links, other than in the most severe and therefore rarest flood events.
49. The EA (EA3/a, para. 9.15-9.19) refer to national benefits of keeping transport links open, and protecting utilities. All of these are capable of inclusion within the BCA but haven’t been, possibly because the EA’s own evidence (CD2.3, Table 8) suggests that the potential damages are too small to warrant inclusion.
50. In any event, when I cross examined Emma Formoy of the EA on 30/11/23, she also accepted there was no evidence that these same benefits would not accrue under A2.
51. The EA also (EA3/a, para. 9.26-9.31) refer to environmental benefits of the proposed scheme, largely around the channel being packaged as a wetland environment. This is a wetland environment that nobody asked for, that risks destroying existing nationally rare grassland, and that won’t reach a natural appearance for many many years.
52. The foregoing all make it impossible to see the EA’s scheme as preferable to an optimised A2 on any sort of economic basis. Overall, omitting the channel makes sense, subject to the validity of the EA’s reasons for insisting on it being retained. I will now consider these reasons.

THE EA’S REASONS FOR WANTING THE CHANNEL

53. The EA reasons for wanting the channel have been summarized in a document (INQ/13b) that you, Madam Inspector, asked the EA to compile. This collates every reference in the EA evidence relating to the need to retain the channel. I have summarised these into four categories^[2].

1.“Uncertainty” and “unreliability”

54.The first category is that of uncertainty and unreliability. The EA suggests that ‘no channel’ alternatives would make the scheme ‘uncertain’ and ‘unreliable’. They make no distinction between these terms – they don’t define or quantify them separately - so I refer to them together.

55. Aspects of uncertainty referred to by the EA include that: the “no channel” option would not increase capacity in the western floodplain, and hence increases risks elsewhere (B); the flood plain does not have not enough capacity to avoid flood impacts (J); the lack of a channel would create a “weak point” where you cannot say with any certainty where that water would go (D); and control of the floodplain would be uncertain without a channel (H).

56. Madam inspector, the EA has provided no evidence on how much increased capacity its scheme would provide. It has provided no quantification of what ‘certainty’ the proposed scheme would provide over no-channel alternatives; no quantification of the additional ‘reliability’ that a channel would provide; or how much additional ‘control’ of the floodplain a channel would provide. It has provided no evidence that the floodplain doesn’t have enough capacity. Therefore any value added by the channel in these areas is unknown and unquantified.

57.The closest the EA gets to providing evidence of certainty is through its hydrological model (CD3.29). And that model shows that Alternative A2 would work nearly as effectively as the proposed scheme. If A2 was optimized it may well be as ‘certain’ as the proposed scheme. Remember that today’s “no channel” floodplain already does a very good, certain job of conveying floodwater to the River Thames.

58. The EA says that flood control structures should not increase flood risk to third party land or it would be liable (G). However the EA has presented no evidence to show that its own scheme, minus one element, the channel, would increase flood risk anywhere. On the contrary, the modelling shows the scheme without the channel increases protection across the area covered by the scheme relative to the current situation.

2. More land to be subject to CPOs

59. Second, the EA claim that their control of the proposed channel would allow them to ensure this preferential route for floodwater is not impeded. They argue that, without the channel, larger areas of floodplain might need to be acquired by CPO to ensure the flow of floodwater (F, Gii, I).

60. It’s hard to see the need for such CPOs when there are other remedies (e.g. agreements with landowners) to ensure obstructions are not created. There is also no evidence that the existing management of the flood plain is in any way deficient, or that this might change. This is hardly surprising as landowners have every incentive to manage their land so as to minimise flooding.
61. Moreover, the EA would struggle to make a case for such CPOs since their modelling indicates that a ‘no channel’ option would work, and they cannot quantify the “reliability / certainty” benefits of having a channel3. Properties protected and net presemnt value
62. Third, the EA claim that the proposed scheme has a better net present value than the alternatives, protects more houses from flooding in severe flood events, and provides more protection from sewer and ground water flooding (A, Gvi).
63. As I already have discussed, differences in NPV and properties protected are largely due to the suboptimal way the alternative have been modelled.
64. Regarding sewer and ground water flooding, I have not been able to find any EA evidence to show the proposed scheme performing better than Alternatives A1 or A2 in this regard.4. Lack of channel could be irreversible
65. Finally, the EA claim that omitting the channel now might not be reversible – it might not be possible to add a channel in the future - due to the requirements of the FCERM guidance for schemes (E).
66. Part of the FCERM guidance (CD5.6) concerns the need to have a positive NPV. If there are doubts over the viability of the channel as an add-on then that implies the channel would not meet this test.
67. If a case could not be made for the channel as a later and separate “add-on” then doesn’t that demonstrate that it adds very little value relative to cost?

CONCLUSION

Given the foregoing here are my conclusions which I hope, Madam Inspector, you will agree with.

The flood protection benefits of the proposed scheme, as summarised in the BCA, are only very slightly more than those of Alternative A2, but there are good grounds to believe that A2 could be optimised could so that its cost could be decreased and benefits increased. The resultant increase in NPV would already make it preferable to the proposed scheme.

On top of this there are the wider costs of the channel: on Hinksey Meadow, the A34, recreation, biodiversity, and the way that cost savings can support other flood measures elsewhere. These make an optimised A2 clearly preferable in terms of wider benefits and cost. Hence it cannot be said that there is a compelling case for the scheme proposed by the EA.

The only ground for retaining the EA’s scheme would be if there were some overriding reason for retaining the channel. The EA have claimed that such reasons exist, but when examined these reasons are wanting.

For all of the EA’s claims of certainty and reliability for a channel, they cannot quantify this is any way.

The importance of certainty depends on the consequences if the system does not behave as expected. Therefore I asked Mr Harding (cross examination 30/11/23) if the EA had looked to see what the consequences might be. His reply was this hadn’t been done.

So rather than trusting their own modelling of A2, the same modelling methodology they used to demonstrate the benefits of their scheme, they instead want the implications of the modelling to be ignored, and their scheme to be adopted in the face of the facts.

The nebulous benefits of the unquantified reliability and certainty claimed for the channel need to be viewed against the certainty of the channel’s very high costs. I cannot see how a public body can justify incurring known costs for unknown benefits.

Madam Inspector, I highlighted at the beginning of my submission that the government guidance on CPOs distinguishes between the intention of the acquiring authority and the wider public interest. The EA’s remit may be limited to reducing flood risks. The EA may have to follow their guidance which refers to all of these costs as only ‘risks’ that can be ‘mitigated’ (CD5.6, Sec. 9.5, page 126; EA3a para. 9.33, p.18). But the public, we, have to face the proposed scheme’s costs as well as its benefits.

The purpose of this inquiry is to act as an external check on what the EA is proposing; to see whether the scheme, with all of its costs, benefits, certainties and uncertainties, is of compelling public interest. The channel costs too much in all kinds of ways. The CPOs for it should not be confirmed.

• ^[1] The bottom of the existing culverts under Willow Walk is 55.0m AOD (ie above mean sea level), which is level with the bottom of the ditch that is parallel to, and in-between, Hinksey Meadow and Willow Walk
• The top of the culverts is 56.1m AOD (diameter 1.1m)
• Hinksey Meadow ground level is an average of 55.7m AOD
• The underside of the proposed bridge is 57.13 AOD (see drawing)

^[2] These show the identifying reference from INQ/13 plus my response.