Stansted Airport Generation 1 Inquiry Rebuttal …stopstanstedexpansion.com/documents/BAA_Air_Quality...BAA/4/D (Case Reference No: 2032278) Stansted Airport Generation 1 Inquiry Rebuttal

BAA/4/D

(Case Reference No: 2032278)

Stansted Airport Generation 1 Inquiry

Rebuttal

PROOF OF EVIDENCE BY Malcolm Pratt

BSc, EurChem, CChem, CEnv, FRSC, FIEMA

In response to UDC/7A, NT/2/A and SSE/7/a

May 2007

BAA Stansted

Rebuttal Proof of Evidence Air Quality – Volume 4

May 2007

Entec UK Limited

Report for ..

..

..

Main Contributors ..

Issued by

…………………………………………………………

..

Approved by

…………………………………………………………

..

Entec UK Limited Windsor House Gadbrook Business Centre Gadbrook Road Northwich Cheshire CW9 7TN England Tel: +44 (0) 1606 354800 Fax: +44 (0) 1606 354810 14899

h:\projects\ea-210\14000 projects\14899 baa stansted\a_post es work\evidence\baa stansted g1 air quality evidence volume 4 rebuttal final.doc

BAA Stansted

Rebuttal Proof of Evidence Air Quality – Volume 4

May 2007

Entec UK Limited

Certificate No. EMS 69090

In accordance with an environmentally responsible approach, this document is printed on recycled paper produced from 100% post-consumer waste, or on ECF (elemental chlorine free) paper

Certificate No. FS 13881

Copyright and Non-Disclosure Notice

The contents and layout of this report are subject to copyright owned by Entec (© Entec UK Limited 2007) save to the extent that copyright has been legally assigned by us to another party or is used by Entec under licence. To the extent that we own the copyright in this report, it may not be copied or used without our prior written agreement for any purpose other than the purpose indicated in this report.

The methodology (if any) contained in this report is provided to you in confidence and must not be disclosed or copied to third parties without the prior written agreement of Entec. Disclosure of that information may constitute an actionable breach of confidence or may otherwise prejudice our commercial interests. Any third party who obtains access to this report by any means will, in any event, be subject to the Third Party Disclaimer set out below.

Third Party Disclaimer

Any disclosure of this report to a third party is subject to this disclaimer. The report was prepared by Entec at the instruction of, and for use by, our client named on the front of the report. It does not in any way constitute advice to any third party who is able to access it by any means. Entec excludes to the fullest extent lawfully permitted all liability whatsoever for any loss or damage howsoever arising from reliance on the contents of this report. We do not however exclude our liability (if any) for personal injury or death resulting from our negligence, for fraud or any other matter in relation to which we cannot legally exclude liability.

Document Revisions

No. Details Date

1 Final Report 21.05.07

BAA Stansted Rebuttal Proof of Evidence Air Quality – Volume 4

Certificate No. EMS 69090

In accordance with an environmentally responsible approach, this document is printed on recycled paper produced from 100% post-consumer waste, or on ECF (elemental chlorine free) paper

Certificate No. FS 13881

i


© Entec UK Limited

14899 May 2007

Contents

1 Introduction 1

2 Mr Moorcroft (UDC/7/A) 2

2.1 Overview 2 2.2 Critical Levels 2 2.2 Critical Loads 5

3 Dr Haycock (NT/2/A) 9

3.1 Overview 9 3.2 Issues Involving the Interpretation of the EU Air Quality

Directive 9 3.3 Issues Involving Air Quality Modelling 10 3.4 Issues Involving Short Term Concentration Variations 14 3.5 Conclusions 14

4 Dr Elliott (SSE/7/A) 15

4.1 Overview 15 4.2 Possible Misunderstandings 15 4.3 Central Issues with Dr Elliott’s Evidence 16

Figure 2.1 (previously Figure 4.2) Annual mean concentration of NOx for the 35mppa case in 2014 modelled using 2003 meteorological data and the north-south and east-west transects 18

Appendix A Response to Detailed Modelling Questions (Haycock Associates 24th April 2007) 19

ii


© Entec UK Limited

14899 May 2007

1


© Entec UK Limited

14899 May 2007

1 Introduction

1.1 In this rebuttal evidence I respond to a number of issues arising from the evidence of Mr Moorcroft (UDC/7/A), Dr Haycock (NT/2/A) and Dr Elliott (SSE/7/A). I have used the headings and sub headings of these witnesses, indicating briefly what point the witness is addressing followed by my response. Some points are raised several times in the same proof or by more than one witness. To avoid repetition I will only address each point once. I have not sought to address each and every point in issue between the parties, and have concentrated my attention on what appear to me to be the more significant issues.

2


© Entec UK Limited

14899 May 2007

2 Mr Moorcroft (UDC/7/A)

2.1 Overview 2.1.1 I note that there is a considerable degree of common ground between Mr Moorcroft and

myself. In addition to the points recorded in the Statement of Common Ground [CD/?], the following points emerge from his proof of evidence:

• He appears to accept our deposition calculations (paragraph 5.2 page12).

• He accepts the nitrogen deposition methodology will over estimate deposition rates (paragraph 5.7 page 14).

• He does not dissent from the interpretation of the application of the NOx limit value for the protection of vegetation but leaves it to legal interpretation. He acknowledges the use of exclusion zones applied by Defra (paragraph 3.8 page 7).

2.1.2 As to the differences between us, I would summarise the key points as follows:

• He differs in his calculation of the rate of decline in deposition rate (paragraph 4.7 page 11].

• He always relates deposition rates to the lowest critical load value (10) (paragraph 4.7 page 11), which he accepts is a worst case assumption (paragraph 5.9 page 15).

• He compares the results from the deposition calculation by mixing the results between the base and sensitivity conditions (paragraph 6.8 and 6.5 pages 16 and 17). In my view that is not an appropriate comparison, for reasons I explain below.

2.1.3 For convenience in the remainder of this section I have grouped the comments on Mr Moorcroft’s evidence under the headings of critical levels and critical loads.

2.2 Critical Levels 2.2.1 In paragraph 3.8 page 7, Mr Moorcroft records that the draft air quality strategy

(CD/186) incorporates the limit value of 30µg/m3 and has interpreted the siting criteria as exclusions zone where the objective does not apply. He goes on to say with regard to the last point that it is a matter of legal interpretation.

2.2.2 I concur with this point. The interpretation I have used in the Environmental Statement (ES - CD/6 paragraphs 11.2.3 to 11.2.7) and in my evidence (BAA/4/C Appendix III paragraphs III.8 to III.11) is consistent with that applied by Defra. I note that this is reflected in the Air Transport White Paper (CD/87 paragraph 11.34), which says:

3


© Entec UK Limited

14899 May 2007

“The NOx concentration limit for the protection of vegetation is not considered to be applicable around a developed Stansted.”

2.2.3 In paragraph 3.9 Mr Moorcroft introduces the World Health Organisation’s (WHO) short term critical level (75µg/m3 as a 24-hour mean), and he goes on to say that the value has not been adopted in the EU or UK legislation.

2.2.4 I have set out a fuller extract from the WHO Guidelines (2001) below because there is a distinction made by WHO in the text between the short term critical level, where it is said that a strong case can be made for the provision of such a level, but at present there are insufficient data to provide such levels with confidence, and a clear recommendation with an unqualified identified level as in the case of the annual mean.

Extract from WHO (2001) page 232.

“A strong case can be made for the provision of critical levels for short-term exposures. There are insufficient data to provide these levels with confidence at present, but current evidence suggests values of about 75µg/m3 for NOx and 270µg/m3 for NH3 as 24-hour means. Interactive effects between NO2 and sulphur dioxide and/or ozone have been reported frequently (8-13). From a review of recent literature, however, it was concluded that the lowest effective levels for NO2 are approximately equal to those for combination effects (although in general, at concentrations near to its effect threshold, NO2 causes growth stimulation if it is the only pollutant, while in combination with sulphur dioxide and/or ozone it results in growth inhibition).

Critical levels for a 1-year period are recommended to cover relatively long-term effects. The critical level for NOx (NO and NO2, added in ppb and expressed as NO2 in µg/m3) is 30µg/m3 as an annual mean. The critical level for NH3 is 8µg/m3 as an annual mean” (my underlining).

The above extract does not read to me with the certainty that Mr Moorcroft (or Dr Haycock NT/2/A paragraph 1.4.3 page 6) suggests. At page 231 it sets out

“Gaps in knowledge”

There have been important developments in the use of critical level and critical load approaches for setting air quality guidelines. With regard to the critical levels of nitrogen-containing air pollutants, however, there are several areas where improvements are urgently required.

• The guidelines for the critical levels of NOx and NH3 are intended to apply to all classes of vegetation and under all environmental conditions. However, more information is needed to quantify the range of sensitivity.

• The guideline for NH3 is based on research performed in temperate climates on a limited range of soil types. To a lesser extent this applies to NOx as well. Caution is required when critical levels are considered for plants in very different conditions, for example in tropical and subtropical zones.

• There is a need to understand further the long-term impacts on growth of changes in biochemical parameters.

4


© Entec UK Limited

14899 May 2007

• There is growing awareness of the physiological importance of NO, and this is reflected in the new incorporation of this compound in the guideline for NOx. Comparisons of the phytotoxicity of NO and NO2 are scarce and still not conclusive with regard to their relative degree of toxicity.

2.2.5 The comments regarding ammonia (NH3) in the above extracts are not relevant to the evidence before this inquiry but I have not deleted these references to ensure a full reproduction of the text.

2.2.6 It is true to say that the WHO guidance on pollutant concentration and their effect is an important source of information used by governments to assist in the establishment of air quality standards. In setting standards however governments have regard to a range of information, including advice from their own national experts and a range of other factors. As Mr Moorcroft states, neither the EU nor the UK has incorporated this short term concentration in legislation. Moreover Defra in its review of the air quality strategy does not propose that this value should be adopted as a policy objective. In fact, I could find no mention of this concentration anywhere in the consultation document (CD/186).

2.2.7 Prior to the publication of the ES (CD/6) there was extensive consultation on the scope of work for the assessment. That process included the production of BAA’s scoping report and UDC’s formal scoping opinion, a workshop with the National Trust, and two meetings with the local authorities on methodology etc. At no time during this process was a request made for an assessment of this short term concentration. After submission of the ES (CD/6), UDC sought a range of further information in its Regulation 19 request (CD/22), including a number of detailed requests concerning the impact of the proposals on air quality. No request was made for an assessment of the short term concentration.

Application of an Exclusion Zone 2.2.8 In paragraphs 3.10 to 3.13 Mr Moorcroft discusses the government’s use of exclusion

zones and its effect on sensitive sites. The issue was recognised in the draft air quality strategy (CD/186), which sets out proposals for a medium term objective as noted by Mr Moorcroft in paragraph 3.11. In paragraph III.10 page 27 - BAA/4/C I have provided a full extract of paragraph 73.

Annual Mean 2.2.9 In paragraph 5.2 Mr Moorcroft reproduces in Table 2 the NOx concentration data that

is also shown in my Table 4.2 (page 3 - BAA/4/C). He observes in paragraph 5.3 that the critical level is attained in 2014 for the 25mppa case (28.8µg/m3) but is exceeded for the 35mppa case (30.6µg/m3). He also observes that the airport makes a substantial contribution to the annual mean concentration (35%).

2.2.10 The figures Mr Moorcroft quotes are common ground, but it is important to understand that the receptor the data is taken from is outside Hatfield Forest, as shown in Figure 4.2 (BAA/4/A p10). Figure 4.2 in my BAA/4/C shows the predicted contours in the 35 mppa case, and it can be seen that the 30µg/m3 contour does not extend into Hatfield Forest itself. In order to illustrate the effect of distance on concentrations, I have produced an additional map, Figure 2.1 which shows the

5


© Entec UK Limited

14899 May 2007

predicted concentrations within the forest area following a north-south transect (553200,221300 to 553200,218600) from the receptor shown in Figure 4.2. The figure shows the fall in total concentration with increasing distance (from the receptor just outside Hatfield Forest used in the deposition calculations) in the 25 and 35mppa cases, together with the percentage contribution of airport related emissions. The concentration in the 35mppa case at the first receptor within the forest is 28µg/m3. The second east/west transect is explained in paragraph 3.3.17 page 11.

2.2.11 The effect of the proposed development at 35mppa is therefore not predicted to result in the critical level being exceeded within Hatfield Forest itself.

2.3 Critical Loads

Approach (2007) 2.3.1 In paragraphs 4.5 to 4.7 Mr Moorcroft sets out the baseline critical load in Hatfield

Forest and Eastend Wood for 2007 using the data from APIS. In Section 5 he goes on to present the equivalent data for 2014. The methodology he uses is the same as I have used (BAA/4/A page 8 paragraph 4.1.16 to 4.1.20). We differ however on how the deposition rates are adjusted to account for the passage of time from the model year (1999-2001) to the present (2007) or future (2014) year.

2.3.2 Both Mr Moorcroft and I cite the same Highway Agency Interim advice note 61/05, an extract from which is included in BAA/4/C Appendix VII. Mr Moorcroft has assumed the reduction is cumulative, whereas the advice note (footnote 1) says “annual decrease in nitrogen deposition can be assumed to be 2% (estimated in a non-cumulative manner i.e. decrease over 5 years is 5x2% = 10%)…..”.

2.3.3 The effect of this different interpretation on the reduction is 13% rather than 14%. The difference is relatively minor. Taking the values in Table 1 (UDC/7/A).

Total N Deposition (kg-N/ha/y) 1999 - 2001

Total N Deposition 2007 (kg-N/ha/y)

at 13%

Total N Deposition 2007 (kg-N/ha/y)

at 14%

Hatfield Forest 36.5 31.7 31.4

Eastend Wood 38.8 33.7 33.4

2.3.4 In paragraph 4.7, Mr Moorcroft sets these values in context against the lower value in the critical range (10kg-N/ha/y), a measure he uses again in considering the significance of the predicted concentrations in the 25mppa and 35mppa cases (paragraph 5.9 and Table 6). In my evidence I use a mid range value of 12kgN/ha/y for the reasons explained in BAA/4/C page 3 Table 4.1 note (b).

2.3.5 In the WHO (2001) Guidance Table 37 (page 250) there is some advice on the selection of the critical load value where ranges exist. This table does not appear to support the selection of the lower value for use in wooded areas of Southern England such as Hatfield Forest and Eastend Wood. However, it does appear to be consistent with the use of 12kg-N/ha/y as adopted by the UK National Focus Group.

6


© Entec UK Limited

14899 May 2007

Critical Loads (in 2014) 2.3.6 In paragraphs 5.4 through to 5.9 Mr Moorcroft sets out the nitrogen deposition data

using the same information as I have used in my evidence (BAA/4/A paragraphs 4.1.21 to 4.1.36). Mr Moorcroft retains the 40mppa sensitivity case data whereas I do not. These data have been omitted from my evidence in view of BAA’s indication that it will accept the imposition of a condition limiting the throughput of the airport to about 35 mppa.

2.3.7 I note that Mr Moorcroft makes no criticism of the nitrogen deposition methodology nor does he offer any alternative methodology.

2.3.8 In paragraph 5.6 Mr Moorcroft states that the “… calculated nitrogen deposition flux is related to the dry deposition of nitrogen dioxide alone, and thus only represents a fraction of the total deposition flux.” As I have explained in my evidence at paragraph 4.1.22 page 10 (BAA/4/A), contributions to nitrogen deposition in the forest that arise from emissions from the airport are due to dry deposition of NO2. Wet and dry deposition of NO and wet deposition of NO2 are generally ignored in long range modelling due to low water solubility or low deposition velocities. These pollutants can however be converted to other chemicals species (e.g. nitrates) over time and then be deposited. The time scale for these reactions are however much longer than the travel time across Hatfield Forest (paragraph 4.1.21 page 10). Ammonia which arises mainly from agricultural and waste disposal activities also makes a significant contribution to nitrogen deposition. It is the contribution from nitrates, ammonia and other similar compounds that is responsible for the fraction that has not been quantified in the total deposition flux. I have mentioned the small contribution of ammonia that arises from airport related road transport (paragraph 4.1.34 page 13) but there are no important airport related contributions that have been omitted.

2.3.9 In paragraph 5.7 (page 14) Mr Moorcroft acknowledges that my assumption (based on proportionately of NO2 to NOx concentrations) will overestimate the airport’s contribution to nitrogen deposition. However, his evidence does not go on to consider the implications of this overestimate. My estimate of the effect of this assumption is about 30% (paragraph 4.1.35, page 13 - BAA/4/A).

2.3.10 There are three main differences in the approach taken to the interpretation of the deposition data between Mr Moorcroft and myself. First, in the rate of reduction in deposition rates (see paragraph 2.3.2 above). Secondly, in the measure used to assess the significance of different critical loads (see paragraph 2.3.4 above). Thirdly, I also believe Mr Moorcroft is incorrect to combine data calculated under different primary NO2 and ozone concentration conditions, as only one set of the conditions can exist in the future.

2.3.11 Taking each point in turn:

a) At paragraph 5.5, the reduction in the deposition rate given by Mr Moorcroft is 25%. This is based on a cumulative assessment, as explained in footnote 2 to his paragraph 4.6. If a non-cumulative approach is adopted, as suggested by the Highways Agency Guidance (BAA/4/C, Appendix VII, p. 53, footnote 1) the reduction in the deposition rate is 28%. The difference in deposition rates in 2014 is therefore:

7


© Entec UK Limited

14899 May 2007

Total N Deposition (kg-N/ha/y) 2014

25%

Total N Deposition (kg-N/ha/y) 2014

28% Hatfield Forest 27.5 26.3

Eastend Wood 29.2 27.9

b) Mr Moorcroft expresses the estimated deposition rates for the 25, 35 and 40mppa cases as a percentage of the lower band of the critical load (Table 6 page 15), which he acknowledges is a worst case assumption. If these percentage are set in context against the mid range value (12 kg-N/ha/y) I have used, the percentage values in this table would be 2 or 3 points lower.

c) In section 5 of his proof of evidence, Mr Moorcroft has set out the calculations for the base and sensitivity cases separately. In section 6, however, in setting out his conclusions, he calculates his percentage increase by mixing the base and sensitivity conditions together (page 16, paragraph 6.9). In my view that is not a valid comparison for the following reasons.

There is uncertainty about the level of primary NO2 that will be emitted from

diesel vehicles in 2014 which was recognised in the ES (page 51 paragraphs 10.4.1 to 10.4.6 – CD/6). There is also uncertainty in the ozone concentrations (which is involved in the conversion of NOx into NO2) in 2014. This was also recognised in the ES (page 51 paragraphs 10.4.1 to 10.4.6 – CD/6) by presenting modelled NOx concentrations at ozone concentrations of 65µg/m3 and 76.6µg/m3. In its Regulation 19 request for further information, UDC asked for the primary NO2 fraction to be increased from the value of 15% used in the sensitivity test to 20%. The results of this sensitivity test were reported at the higher ozone concentration in the Regulation 19 Response (CD/22) as set out in my evidence (paragraph 4.1.26 page 11 – BAA/4/A). These uncertainties will affect both the 25mppa and 35mppa cases, and only one primary NO2 fraction and ozone concentration can exist in the Stansted area in 2014. As a consequence of this, Mr Moorcroft should compare the 25mppa and 35mppa cases under either the base or sensitivity test conditions and not mix them.

2.3.12 It follows from this that Mr. Moorcroft’s figure of 45% for 35 mppa (page 16, paragraph 6.8) cannot be correct. The correct comparison should be

Estimated Airport Contribution to N Deposition flux (kg-N/ha/y)

mppa Base Sensitivity

Hatfield Forest 25 1.1 1.3

35 1.3 1.6

Difference 0.2 0.3

Percentage increase 18 23

8


© Entec UK Limited

14899 May 2007

Estimated Airport Contribution to N Deposition flux (kg-N/ha/y)


Eastend Wood 25 1.2 1.4

35 1.5 1.8

Difference 0.3 0.4

Percentage increase 25 27

Similarly in Table 5 (page 15 – UDC/7/A) you cannot compare the contribution to the total nitrogen deposition of 4% (25mppa base) to 5.8% for the 35mppa sensitivity case as set out in paragraph 6.9. The correct comparison is set out below (from Table 5).

Airport Contribution to N deposition (%)


Hatfield Forest 25 4.0 4.7

35 4.7 5.8

Difference 0.7 1.1

Eastend Wood 25 4.1 4.8

35 5.1 6.2

Difference 1.0 1.4

2.3.13 In paragraph 6.9, the 16% figure is taken from Table 6 and is the percentage of the lower critical load value for the sensitivity conditions for the 35mppa case. This would be 13% under the base conditions. The equivalent figures for the 25mppa case are 13% and 11% so the difference due to the development is an increase of 3% (16 - 13) under the sensitivity conditions or 2% (13 – 11) under the base conditions.

9


© Entec UK Limited

14899 May 2007

3 Dr Haycock (NT/2/A)

3.1 Overview 3.1.1 Dr Haycock has taken quite a different approach to that adopted by Mr. Moorcroft in a

number of areas.

3.1.2 I note that in respect of the deposition of nitrogen, Dr Haycock acknowledges that this is part of the general poor situation in southern England, and a regional problem. For this reason the National Trust’s evidence does not focus on the deposition of nitrogen. That approach is consistent with the views expressed by the National Trust in response to consultation about the issues that should be addressed in the ES (ES Volume 3, page 3, Table 1, CD/6). In my view that approach is correct.

3.1.3 Dr Haycock does not challenge any of the work on nitrogen deposition (except indirectly by saying NOx (and by implication NO2) concentrations have been underestimated because the emissions are too low and the spatial extent is underestimated. I believe that these concerns are not well founded, for the reasons I set out below.

3.1.4 Dr Haycock’s concerns seem to reduce down to the following:

1. The application of the NOx air quality limit value, and in particular a concern that Defra has misunderstood the effect of the relevant EU legislation.

2. The choice of model made by BAA, and whether a different model might have been better.

3. The validation of the model used, and the reliability of the predictions made.

4. The short-term concentrations of NOx that are, and are likely to be, experienced in Hatfield Forest.

3.2 Issues Involving the Interpretation of the EU Air Quality Directive

3.2.1 Dr Haycock and I differ on our understanding of the application of the NOx limit value for the protection of vegetation. As I have said earlier in this rebuttal evidence (paragraph 2.2.2) I have followed the interpretation of the application of the limit value given by Defra as I have no legal expertise. I note that Dr Haycock does not accept the Defra interpretation of the directive.

10


© Entec UK Limited

14899 May 2007

3.3 Issues Involving Air Quality Modelling

Atmospheric Dispersion Modelling 3.3.1 Dr Haycock’s position on the modelling undertaken and reported in the ES (CD/6)

appears to be that I have used the wrong model, omitted some emissions and there is inadequate spatial representation of the concentration arising from the emissions. Many of the detailed points of concern are set out in his list of questions (NT/2/C Appendix 4). My responses to these questions are set out in Appendix A to this rebuttal.

3.3.2 I do not accept that I have used the wrong model, that any significant emissions from the airport have been omitted, or that there is an inadequate spatial representation of the concentrations arising from these emissions. I would also note that many of the emissions that Dr Haycock believes are omitted have been identified in the model test (CD/189) and the methodology (CD/190) reports which is a reflection of the diligence of my team in the execution of the modelling work.

3.3.3 In my evidence (Section 3 and Sections 5.1 and 5.2 together with the supporting Appendices) I have explained some of the issues of concern to Dr Haycock as they are common themes in some of the evidence of others.

3.3.4 Before dealing with some specific points raised by Dr Haycock in his evidence, I would like to contrast his view of the modelling with that of Mr Moorcroft (a team member who assisted in the preparation of the Bureau Veritas reports (CD/144 and 145)). In my evidence (BAA/4/A paragraph 3.1.6 page 5) I set out the view of Bureau Veritas, which para-phasing said, the air quality report is thorough overall, they identify an uncertainty in the verification (which I have discussed elsewhere (Appendix II paragraph II.15 (page 15) to II.42 (page 21) - BAA/4/C) and despite the verification concern it is not considered likely that the health based objectives would be exceeded based on the model predictions and experience at other airports.

3.3.5 There is no criticism of the model itself, the emission inventory or the spatial resolution of the model with respect to any pollutant. Bureau Veritas made a single verification point which relates to the comparison of model predictions against measured values.

3.3.6 Returning to the model selection. ADMS is a suite of models that are designed to be used for differing applications. Much of the detail within the models however is common. Each ADMS model can be used for a number of applications as I have illustrated in Appendix II paragraph II.45 (page 21) to II.50 (page 22) BAA/4/C).

3.3.7 As a consequence of this, the spatial distribution of concentrations arising from emissions will be essentially the same within each of the models provided that a like for like comparison is undertaken.

3.3.8 The ADMS suite of models is designed to calculate hour by hour concentrations at each receptor from all emissions under defined meteorological conditions. At the end of this process the hourly concentrations are combined to give the annual average concentration at each receptor. The kernel version of ADMS developed by AEA uses exactly the same dispersion equations to calculate the annual average concentration, with the same emission data and meteorological data. The only difference between the kernel and the standard ADMS relates to the way the annual average calculation is

11


© Entec UK Limited

14899 May 2007

undertaken. The spatial distribution of the concentration arising from the emission will be the same for both models if the model parameters, spatial resolution of emissions, and distribution of receptors are common.

3.3.9 In view of this, the spatial distribution of NOx cannot be poorly represented as suggested by Dr Haycock (paragraph 3.8.1 page 21)

3.3.10 This leaves the question of whether another of the ADMS models (ADMS Roads or ADMS Urban) would have been a better choice. The answer in my opinion would be no. Both of these models contain refinements that will tend to reduce concentrations when compared to the standard (or kernel) version of ADMS. For example, both versions take account of the additional vehicle induced turbulence within the road element of the models. The effect of this additional turbulence will be to reduce the near road concentrations compared to that given by the standard (or the kernel version) of ADMS which does not.

3.3.11 In my opinion the kernel version of ADMS produces slightly higher concentrations than either ADMS Roads or ADMS Urban and hence the approach I have used is more conservative.

3.3.12 The kernel version of ADMS does have a much shorter run time than the other versions of ADMS but this was not the reason for its choice. The reduced running times stems from removing the requirement for the calculation of the hour by hour concentration as the focus of the study was the annual average rather than short period concentrations.

3.3.13 At paragraph 3.3.3 page 18 Dr Haycock suggests “the kernel approach … represents a first approximation to determine gross impacts on human health”. This is not correct, for the reasons I have given above.

3.3.14 ADMS Roads was available but was not used for the reasons given above. Moreover, if this model was used in the hour by hour mode, the limited number of point and area sources available within the code would have limited the size of the study area that could be covered in a single run.

3.3.16 Dr Haycock (paragraph 2.1.1 page 8 – NT/2/A) notes that “… high NOx concentrations associated with road networks, degrade quickly the further one moves from the centreline of the carriageway (often within less than 500m) ...”. Similarly, Mr Moorcroft (paragraph 3.13 page 8 – UDC/7/A) states that “pollutant concentrations decline rapidly with increasing distance from the road, such that levels become imperceptible from the general background within about 200m”. He goes on to say “the contribution from road traffic on the M11 to concentrations of nitrogen oxides at Hatfield Forest would be very small and probably immeasurable”.

3.3.17 In Figure 2.1, I have also provided the NOx concentrations along an east/west transect running across the M11 for the OS co-ordinates 550000,220000 to 553200,220000. This figure shows the total NOx concentrations (from all sources) together with airport related and non-airport related road traffic for the 35mppa case. While it is evident that concentrations do fall off rapidly within the distances referred to by Dr Haycock (and Mr Moorcroft) the contribution to concentrations at the edge of Hatfield Forest is about 4µg/m3 or 13% of the limit value from non-airport related vehicles alone. If the airport related vehicles are included the contribution would increase to 15%. I would not consider this to be immeasurable.

12


© Entec UK Limited

14899 May 2007

3.3.18 It is important to understand that modelled concentrations from any development in a current or future year are always a best estimate. There is inevitable uncertainty in the calculated concentrations due to the model itself but there are probably much greater uncertainties in the forecasting of emissions in future from all the various sources.

3.3.19 The modelled concentration comprises the best estimate of what the effects might be in the future in the 25mppa and 35mppa cases. In the comparison between the cases, all the uncertainties in the model and the emission data remain the same for both cases and hence the modelling of these differences is robust. Table 3.1 (page 1 - BAA/4/C) illustrates the differences in concentrations at specified receptors between the 25mppa and 35mppa cases which are generally small.

Model Validation 3.3.20 Model validation is a matter for the developers of the model and its computer code

(Appendix II paragraph II.15 page 15). The criticism made by Dr Haycock (and others) relates to the on-site verification. In particular his concern focuses on the single site and the 7 month measurement period (paragraph 3.4 1 page 18).

3.3.21 I have set out my opinion on these matters in my evidence at Appendix II paragraph II.15 (page 15 to II.42 page 21 - BAA/4/C). The verification study used the available monitoring data at Stansted coupled with other equivalent studies undertaken at Gatwick and Heathrow using the same modelling methodology.

3.3.22 Dr Haycock is critical of the scope of the monitoring programme put in place following the 2003 consent as set out in the s106 agreement. As I understand it, BAA has installed the monitoring equipment to accord with the s106 agreement. The location of the monitoring equipment has also been agreed with UDC. Dr Haycock contends (paragraph 3.5.3 page 19) that “poor implementation of this monitoring network has ….. critically undermined the testing of the model in the ES”. This is not accepted for two reasons. Firstly, my understanding is that BAA has complied with the requirements of the s106, and it is not correct to say the monitoring network has been poorly implemented. Secondly, the verification work at Stansted, Heathrow and Gatwick shows the model performs well and provides a robust assessment of the likely effects arising from the proposed G1 development (paragraph 5.25 page 17 - BAA/4/A).

3.3.23 I am not aware of any issue of poor maintenance that affects the BAA data (except in relation to carbon monoxide concentrations as noted in Table 10 page 25 of the ES (CD/6)). Data capture for NOx, NO2 and PM10 is better than 90%.

3.3.24 At page 19 paragraph 3.5.1 Dr Haycock discusses diffusion tubes, and refers (line 11) to NOx which might be incorrect. To avoid doubt, diffusion tubes used at Stansted measure NO2 and as stated provide information on spatial distribution of concentrations. If a diffusion tube is co-located with a continuous NOx/NO2 analyser, then the NO2 concentrations measured using tubes elsewhere (in similar exposure environments) can be corrected against the reference measurement. This is called “bias” correction. Where a co-located tube is not available then correction factors can be taken from other co-location measurements undertaken by others using the same tube type, analytical laboratory and locations (paragraph II.11, page 14 BAA/4/C).

13


© Entec UK Limited

14899 May 2007

3.3.25 An NO2 diffusion tube cannot be simply used to calculate NOx concentrations (except perhaps at low NO2 concentrations) because of the non-linear relationship between NOx and NO2.

3.3.26 It was expected that the UDC site at Takeley would have been used as a second site for the model test work but as set out in the report (CD/189 – page 49 paragraph 4.2.5) doubts were raised about the quality of the data in the first few months of 2004 and for this reason these data were not used.

3.3.27 Co-located diffusion tubes at High House would not have provided any further information on NOx.

3.3.28 The diffusion tubes were used in the model test work (paragraph II.36, page 20 - BAA/4/C).

3.3.29 In my view there was no need for more monitoring data, as the performance of the model has been tested elsewhere and there is nothing unique about Stansted that prevents these data from being used.

Dr Haycock’s List of Questions 3.3.30 The list of questions drafted by Dr Haycock raises a number of related issues:

• On the period and data inputs for the model test (verification) study;

• The suitability of the model for the study;

• The selection of model parameters and their sensitivity to the reported concentrations;

• Entrained effects (of NOx in aircraft vortices) not being included in the modelled output;

• Request for detailed model output files or additional contours intervals; and

• Quantification of the cumulative error arising from emissions sources that were discounted.

3.3.31 Following a meeting with Dr Haycock (26th April) I have provided a written response to Dr Haycock, which is included as Appendix A to this rebuttal evidence.

3.3.32 The questions raised and the responses I have provided do not have a significant effect on the modelled concentrations presented in the ES (CD/6). The model used provides the best estimate of future concentrations. My opinion on the suitability of the modelling approach used for this assessment has not been changed by my discussions with Dr Haycock, or consideration of the questions he raised. I consider that the modelled concentrations before this inquiry provide a robust assessment of the likely effect of the development.

14


© Entec UK Limited

14899 May 2007

3.4 Issues Involving Short Term Concentration Variations

3.4.1 Dr Haycock accepts there is no formal short term (24-hour average) NOx limit for vegetation (paragraph 4.2.3 page 23), but he believes the assessment presented in the ES (CD/6) should have addressed this metric. I have already explained (paragraph 2.2.4 page 3 and paragraph 2.2.7 page 4) the reasons why this metric was not addressed, and I do not accept the absence of the 24-hour metric has any significance to the conclusions of the assessment set out in the ES (CD/6) or my evidence. Even if the metric was provided I do not believe it would assist in the determination of this application.

3.4.2 Dr Haycock sets out some monitoring data in his evidence (Figure 2 and Appendix 5) in relation to the running 24-hour concentration. At the time of writing I have not had the opportunity to look at the detail behind these data but I hope by the time this evidence is presented I will have the opportunity to discuss it with Dr Haycock in order to clarify and hopefully agree various technical matters.

3.5 Conclusions 3.5.1 Dr Haycock has raised a number of detailed points arising from the assessment set out

in the ES (CD/6). For the reason I have set out in my evidence (BAA/4/A) and in this rebuttal none of the points raised have a significant effect on the magnitude of the model concentrations I have produced. Other models could have been used but from my understanding of the various refinements within them they are likely to produce lower concentrations than set out in the ES (CD/6). I am satisfied therefore that I have produced a robust assessment of the likely effects on air quality of the proposed development at 35mppa compared to the base case (25mppa) in 2014.

15


© Entec UK Limited

14899 May 2007

4 Dr Elliott (SSE/7/A)

4.1 Overview 4.1.1 Dr Elliott takes a broad view in her evidence and covers some of the same ground as Dr

Haycock and Mr Moorcroft. The main thrust however appears to be centred on:

1. reduction in concentrations (from the 2003 consent) ES at 25mppa compared to those in the current application;

2. air quality assessment and modelling (including proposed changes arising from the PSDH work); and

3. the effect of the expansion.

4.1.2 I will address these themes below. First however I would like to address a few matters which might have been misunderstood.

4.2 Possible Misunderstandings 4.2.1 In paragraph 3.2.3 (page 3) Dr Elliott notes that “These (diffusion tubes), BAA claims,

have been calibrated by comparing them with automatic monitors...”. Co-located diffusion tubes were used as part of the 3-month monitoring programme (September 2001 to January 2002) as reported in the ES (paragraph 8.1.32 page 26 (CD/6)), but these data have not been used to “calibrate” any diffusion tubes. In the next paragraph (8.1.33) of the ES it says that the NO2 concentrations in Table 11 (page 26) are uncorrected and likely to be overestimates and should be viewed with caution. The monitoring data to which Dr Elliott refers was collected to satisfy the s106 agreement (as noted above in paragraph 3.3.22 page 12).

4.2.2 In paragraph 3.3.7 (page 7) Dr Elliott refers to the claim that background contributions to the PM10 concentration are 95% of the total and relates this to Table 21 (page 37 – CD/6) from which she derives the percentages of 18% at 25mppa and 23% at 35mppa. She then goes on to say the 95% figure cannot be correct. I am not entirely sure where the 18% and 23% figures come from. In any event Dr Elliott appears to be comparing concentration data (Table 24 page 41) with emission data in Table 21 (page 37). The airport related PM10 (PM2.5) emissions are shown (page 38) to increase by 21% from 25mppa to 35mppa. At Table 24 the PM10 background concentrations at High House is given as 18.7µg/m3 compared to a total concentration of 20.4µg/m3; the background fraction at this site is therefore 92% and at the Takeley site it is 96% etc.

16


© Entec UK Limited

14899 May 2007

4.3 Central Issues with Dr Elliott’s Evidence

Reduction in emissions/concentrations since the 2003 consent 4.3.1 Dr Elliott queries the reduction in concentrations since the previous ES (CD/28). I have

explained the reasons for this in my evidence (paragraph 5.2.9 page 18 BAA/4/A) with some more detail at paragraph II.51 (Appendix II, BAA/4/c) and in the model test report (CD/189 page 57). As I said in my evidence, these changes simply follow the evolution in the methodology, which is still occurring through the PSDH programme. I would expect emission rates and concentrations generally to fall through this evolutionary process as better data and methodology will remove some of the conservative assumptions that are inevitably introduced to ensure concentrations are not underestimated.

Monitoring and modelling 4.3.2 Dr Elliott raises many areas of uncertainty that cover for example monitoring, proximity

of the M11, the PSDH work and its effect on emissions and concentrations, data used within the model such as fleet mix, thrust settings, road traffic flows, primary NO2 emissions from diesel vehicles; new modelling techniques arising out of the PSDH work etc.

4.3.3 While there is acknowledgement in the ES (section 6 page 11-CD/6) that there are limitations and assumptions within the assessment, my team used the best data that are available from Stansted and elsewhere to provide a robust assessment of the likely effect of the proposed development. Where doubt exists on the quality of the data our assumptions have always tended to be conservative to provide what I would call “realistic worst case estimates”

4.3.4 The sensitivity test that examines the change in the primary NO2 fraction from diesel vehicles (ES section 10.4, page 51) was undertaken to address matters such as higher NO2 emissions from particle traps raised by Dr Elliott (paragraph 3.2.12 page 4). This test was extended at the request of UDC and reported in the Regulation 19 response (CD/22 at pages 26 to 28). The Bureau Veritas report CD/144 (paragraph 5.1 page 11) acknowledges that this test addresses such emissions.

4.3.5 Dr Elliott identifies a range of issues that have arisen from the PSDH work which was published in July 2006, some 3 months after the publication of the ES. Dr Elliott is correct to say there are uncertainties in the quantification of emissions and the modelling of these emissions. In most cases these uncertainties tend to lead to an overestimate of concentrations which in the Heathrow area is more likely to suggest a breach of an air quality objective when it might not exist. The exceedance being due to the high background concentration and a conservatively modelled airport and road contribution.

4.3.6 In relation to this concern, the work in the ES (CD/6) was specifically reviewed by Bureau Veritas against the main findings given in the PSDH report. The question posed by UDC (page 16 – CD/144) was

“To what extent has the recent work as part of the Sustainable Heathrow project to refine AQ modelling been reflected in the approach at Stansted?”

4.3.7 The conclusion reached by Bureau Veritas was;

17


© Entec UK Limited

14899 May 2007

“In general terms, most of the critical issues identified by the PSDH project have been incorporated into the approach used at Stansted. Importantly, consideration has been given to the near-field dispersion from jet engines, realistic take-off thrust settings have been applied, and the approach used to estimate NOx:NO2 transformation allows the potential impact of increasing primary NO2 emissions and ozone background to be accounted for.”

4.3.8 The AEA modelling approach referred to as parameterising (paragraph 4.2.12 page 13) the effect of jet plume rather than including its effect explicitly can be important when the concentrations of interest are very close (say a few hundred metres) to the source (often referred to as the near field). As distance increases however the differences between the two approaches diminishes. As noted elsewhere in my evidence (Appendix II paragraph II.26 page 18 (BAA/4/C)) the AEA approach tends to overestimate the contribution from aircraft in the near field.

The effect of the expansion 4.3.9 Dr Elliott examines the effects of the proposed development in the context of the

current baseline (paragraph 3.3.2 page 6) and from the data in modelling Tables 21 to 28 (CD/6). The modelling data are compared directly with the relevant objectives in both the 25mppa and 35mppa cases. The difference between the cases is also shown. I have tried to put the change in concentrations into context by modifying the NSCA guidance on significance. This context can be used or disregarded as Dr Elliott has done. The information in the assessment however still shows that all the health based objectives would be achieved and the NOx limit value for vegetation protection is not exceeded within Hatfield Forest or Eastend Wood.

4.3.10 While Dr Elliott might like the PM2.5 standard that is adopted in the US to apply in England (paragraph 3.3.8 page 7), it does not. The nearest proposal to this we have in Europe is a concentration cap of 25µg/m3 as identified in the ES (CD/6, Table 4, page 16). This proposal would be met if it is introduced as set out in the proposed directive (page 133, Appendix XIX (BAA/4/C)).

18

h:\projects\ea-210\14000 projects\14899 baa stansted\a_post es work\evidence\baa stansted g1 air quality evidence volume 4 rebuttal final.doc © Entec UK Limited 14899 May 2007

Figure 2.1 (previously 4.2) Annual mean concentration of NOx for the 35mppa case in 2014 modelled using 2003 meteorological data and the north-south and east-west transects

In electronic copies of this rebuttal this figure is a separate file “Figure 2.1 NOx 35mppa with SSSI and transects.pdf”

19


© Entec UK Limited

14899 May 2007

Appendix A Response to Detailed Modelling Questions (Haycock Associates 24th April 2007) 8 Pages

The responses set out below were discussed with the National Trust and its advisors (Haycock Associates) and BAA and its advisors (Entec UK Limited) at a meeting held on 26th April in Entec’s London office. Each response is indented and follows the question raised by Haycock Associates.

Notes from the BAA model test report (CD/189)

E4 The air quality dispersion model was only validated at one site (High House) for a period of 7 months. Are there now 12 months of data that could be used for model validation? If so, has the model been revalidated? If not, why not?

The model test work was undertaken using monitoring data recorded at High House during the period October 2003 and May 2004. After this period monitoring terminated (June 2004) and recommenced in November 2004 and has continued thereafter. The seven month data period was used for the model test work to assess the model performance to inform the EIA process. As the original test covered operational conditions at the airport that are reasonably uniform throughout the year and the range of meteorological conditions that can be expected over the year there is no reason to expect a longer period would have given a significantly different outcome. In view of this, there was no justification to repeat this substantial piece of work.

E8 Aircraft at ground-level are acknowledged as a more significant source of NOx than airside emissions from vehicles, car parks and heating plants. Benzene can be used as a tracer for aircraft emissions by measuring its ratio to NO2 to aid source apportionment studies (this technique has been used at Heathrow). Why was this not carried out at Stansted?

Haycock Associates to provide a reference to this work (not received as at 21st May 2007).

E14 At High House 57% of NOx is estimated to be from explicitly modelled sources with 45% of this from airport sources. This is therefore concluded to be a good site for testing the model. Why was the model only validated against measurements at one site? Why was validation not also carried out for a site that would be judged to be less affected by airport emissions?

The monitoring station was established at High House as the 15+ modelling work suggested the NO2 objective (40µg/m3) might be exceeded in the area close to the airport. Subsequent monitoring at this site however has proved that this objective has not been approached let alone exceeded. The 15+ work indicated that emissions from the airport should be identifiable against the background due to other sources at this location. The model tests confirm the airport contribution at 57% and hence it was judged to be a good site for the

20


© Entec UK Limited

14899 May 2007

model test. No other continuous monitoring sites were available to explore the airport contributions to concentrations (diffusion tubes were used to examine NO2 concentrations and more distant sites were used to examine the background contribution methodology).

If more distant sites in the Stansted area had been available and used for this test, they would be subject to smaller contributions from the airport (the contribution decreasing with increasing distance). At such locations the background contributions would become dominant and it would become more difficult to see the contribution from the airport. It should be remembered that the purpose of this integral test was to examine the complete modelling process (i.e. emissions from the airport and their dispersion). The performance of the background model in the Stansted region was examined and reported in the model test report (CD/189).

E20 What evidence is there that the road emissions are over estimated?

The modelling process used for G1 does not include the additional dispersion due to vehicle induced turbulence that is included in, for example ADMS Roads. Figure 1 (below) has been reproduced from data included in Figure 4.4 (page 212)1 but includes the identity of the monitoring sites.

Note the position of the M25 site. If the allowance is made for the vehicle induced turbulence then the sites close to roads (M25, Hillingdon and LHR2) all show a reduction in the NOx concentration.

Figure 1 Comparison of modelled and measured NOx concentrations at Heathrow showing the concentration reductions when vehicle induced turbulence is taken in to account

1 Dft 2006. Project for the Sustainable Development of Heathrow Report of the Airport Air

Quality Technical Panel. July

21


© Entec UK Limited

14899 May 2007

0

50

100

150

200

250

300

350

0 50 100 150 200 250 300 350

measured

mod

elle

d Revised MICMotorway sensitivity

M25

HillingdonLHR2

Main Road

Oaks Road

Colnbrook

Hounslow 2Green Gates

Key: “Revised MIC” refers to a Model Inter-Comparison study that formed part of the Project for the Sustainable Development of Heathrow (PSDH)

2.2.34 “There is a tacit assumption that aircraft times-in-mode do not have a significant seasonal variation”. Why was no data collected for winter months for model validation purposes? Are times-in-mode data available from other airports and if so does it support the assertion of no seasonal variation?

The time-in-mode data were collected by NATS. For Heathrow and Gatwick airports we have a large data sample for both summer and winter periods, and a significant systematic effect of season is not detectable against the background intrinsic variability in these parameters. Operations at Stansted are reasonably uniform throughout the year and therefore a significant seasonal effect would not be expected.

2.7.2 Traffic data used in the model test was estimated rather than derived from new road-modelling output. This was considered acceptable because High House was not close to a major road. What validation was carried out for the new road-modelling data once available? Does it affect the reliability of this model test that the road data was derived in a different way?

There was no subsequent validation of the 2003/2004 traffic data used in the model test work against the 2003 data. As previously noted the purpose of the model test was to examine the performance of the overall methodology (emissions and dispersion) at High House (close to the airport but not close to a major road) the level of accuracy and detail in representing road vehicles need only reflect this requirement. Hence it did not affect the reliability of the model test.

22


© Entec UK Limited

14899 May 2007

2.10.7 The ratio of NOx from airport/non-airport related traffic is sensitive to the size of the road network. On the 45 km E-W by 43 km N-S road network, the ratio is 10%. What is the ratio on the network closer to the airport, e.g. on the modelled 10 x 12 km grid?

The ratio of NOx from airport and non-airport related traffic from within the 10x12km study area is about 30%.

2.10.13 It is stated that more detailed information on reverse-thrust use at Stansted would be beneficial. What use has been made of any extra data gained since producing the model test report?

No additional reverse thrust data have been obtained so the question on how it might be used did not arise.

4.1.2 Why was the decision made to use empirical conversion factors for the conversion of NO to NO2 rather than the ADMS chemistry module? Was this because the model module was based on 12 months of data but only 7 months of measurements were available?

To use the chemistry module in ADMS requires the model to be used to calculate hourly concentrations. As the modelling was undertaken using the “kernel” approach (CD/189 page 40) this option was not available. The calculation of NO2 concentrations outside of a dispersion model is a common and well established practice so there was no requirement to use the chemistry module within ADMS.

4.1.2 The section says that “interest lies in off-airport concentrations”. Why therefore was High House selected as the test point as it is very close to the airport?

See response to E14.

4.1.3 There is stated uncertainty on the most appropriate value to use for the lower limit on the M-O length2 and the ADMS default albedo3 of 0.23 was used. Was any work done to look at the effect of the input parameters on modelled dispersion? What account was made of the variations in albedo across the modelled area, for example such as the albedo gradient change between adjacent road and forest areas?

The work undertaken to select the Monin-Obukhov length was explained in paragraphs 4.5.14 to 4.5.17 of the model test report (CD/189). Beyond this work we relied on sensitivity work undertaken by CERC (http://www.cerc.co.uk/software/publications.htm) that showed that adjusting the Monin-Obukhov length within a realistic range had little effect on concentrations. No specific work was undertaken to examine the effect of changing the default albedo value on modelled concentrations. The model

2 Monin-Obukhov length is the distance scale from the surface at which buoyancy effects and

shear effects become comparable. It is used to set a constraint on how 'stable' the atmosphere can become (where stable conditions inhibit the vigour of the turbulence responsible for atmospheric diffusion).

3 surface albedo is the ratio of the reflected to incident short wave radiation at the surface of the earth

23


© Entec UK Limited

14899 May 2007

allows the use of a single albedo value and therefore there was no opportunity to consider a gradient range,

4.1.12 Modelling seems to emphasise the worst case NOx concentrations for on-airport locations. What is the worst case for ambient air if there is more dispersion away from the airport?

Dr Haycock to reconsider the question but as at 21st May no reply has been received.

4.1.17 Known entrainment effects were not included in the modelling on the basis that they would not contribute significantly to annual mean NOx concentrations and that the area of significant impact was unlikely to coincide with that from ground level sources. What modelling work was carried out to support this decision?

None – we relied on the work of Graham and Raper (2003)4 to set the scale of the potential impact.

4.1.18 A similar approach to that used to model roads in the test report was thought to overestimate concentrations close to roads, either because of overestimation of emissions or underestimation of dispersion. As a result, “modelled concentrations very close to roads are correspondingly uncertain”. What distance is meant by very close?

The closest receptor to any road was 10m.

The uncertainty in the “near roads” concentration is considered to exist for tens, not hundreds of metres.

4.1.18 Given the kernel, additive modelling approach adopted why was a specific road model (e.g. ADMS-Roads) not used?

ADMS3.2, which was considered to overestimate near road concentrations, was effectively used as a screening tool. If concentrations greater than 40µg/m3 NO2 had been identified more detailed modelling using ADMS Roads would have been undertaken. The initial modelling did not identify any properties at risk of exceeding this concentration and therefore no more detailed modelling was required.

4.2.8 No NO2 tubes were co-located with the continuous analyser. Why was this, considering such co-location is standard practice?

The BAA monitoring programme was designed to meet the s106 requirements which did not include co-located diffusion tubes at the continuous monitoring site. Monitoring was already in progress before the model test work was commissioned.

4 Graham A and Raper D (2003) Air quality in airport approaches: impact of emissions entrained by vortices

in aircraft wakes. Research report. Centre for Air Transport and the Environment, Manchester Metropolitan University.

24


© Entec UK Limited

14899 May 2007

4.3.7 Was Heathrow the only suitable meteorological data available? Has/is it possible to repeat the analysis with more local data given that the use of this data is then referred to as one of “a number of approximations”.

As explained in paragraph 4.3.1, the background concentrations were derived from a pre-existing methodology that is used to generate national background maps. The choice of met data is an intrinsic part of that methodology. It should be noted that the impact of any approximations is constrained (and then quantified) via the adjustment of model results to national monitoring data.

4.3.9 The prediction errors in the background modelling (+29%, -15%) are within the range of the model measurement differences found in data used to verify the background model. Has any subsequent modelling been undertaken to limit these errors, and if so, what factors contributed to the two points closely located to each other being at different ends of the error scale?

This statement is based on the verification tests of the national background model carried out by authors of that model, which shows that most modelled values lie within a +/30% of the measured value. This is representative of the intrinsic precision of the national background model. The model was considered fit for purpose by its developers, and this level of precision has been accepted in using it to predict background concentrations around Stansted. No evidence is presented in the verification to suggest that the scatter is significantly less if the data points are restricted to a smaller geographical area (say of order tens of kilometres). For example, the scatter does not appear much less for sites within London.

4.4.3 NO2/NOx ratios are based on data for a 7 month period. How might the relationships be affected by the seasonal O3 cycle? The final modelling methodology is based on a 12 month period and a different method of looking at the relationship is used. How have the two methods been compared?

The primary comparison in the model test report relates to NOx concentrations – this tests the emissions methodology and the dispersion modelling. The performance of the NOx to NO2 relationship (for annual means) has been tested separately in many studies, including studies around airports. It is invoked here primarily to make some use of the diffusion-tube data, which is anyway subject to greater uncertainties. The use of a 12-month relationship with a 7-month concentration is an approximation, but the approximation has at least been tested at High House. As stated in the report, no other work was done on the potential difference between 7-month and 12-month relationships. With reference to the change of methodology for the future cases, comparisons were indeed carried out and reported in CD/190 (paragraph 4.3.23 and Table 4.3).

Notes from The BAA Model Methodology Report (CD/190)

E2 Could maps be supplied of current annual mean NOx concentrations for the three modelled meteorological years (2001-2003)? We would like a thorough understanding of the current air quality situation in order to attain an informed judgment of the impact of the proposed development.

25


© Entec UK Limited

14899 May 2007

There are no modelled concentration data for 2001, 2002 and 2003. Meteorological data from these years were used to model emissions from the 25 and 35mppa cases in 2014.

BAA will commission the additional work to produce the contours from the model test study

E3 The model calculates annual mean NOx concentrations. Shorter-term metrics are then derived from empirical relationships rather than through modelling. What difference is there between the shorter-term mean concentrations derived this way and those derived by running the model specifically to gain short-term means?

As noted above (4.1.2) the ADMS based “kernel” approach cannot be used to derive hourly concentrations so comparisons cannot be made. If the period NO2 concentration (28.8µg/m3) is used to examine whether the peak hour concentration objective (200µg/m3) is exceeded by applying the approach used in the ES (CD/6 page 47) the 99.8th percentile concentration would be estimated at 110µg/m3 (107µg/m3 based on the modelled period concentration) compared to the maximum measured concentration of 113µg/m3. Note that the 99.8th percentile concentration corresponds to the 18th highest value which would equate to an exceedance of the maximum hour objective. This comparison suggests the empirical relationship provides a good estimate of this short period metric.

4.1.52 Can we obtain a copy of the air quality models grid of receptors which was used to produce the annual mean NOx contours for the 12 x 10 km modelled area. Ideally, if the grid receptor locations could be sent as an electronic OS coordinated XY text document it would be most appreciated. In addition, could mean annual NOx concentrations for the modelled metrological years (2001-2003) be supplied, plus modelled mean annual NOx concentrations for 2014 based on the airport usage of 25 and 35 mppa.

Dr Haycock was asked to explain why these data were required (email 9th May 2007) as at 21st May no reply has been received.

4.1.53(1) The annual mean NOx concentration contours were produced using triangulation with linear interpolation. Why was this method selected? Were any other interpolation techniques used to check differences in the results obtained?

The resolution of the receptor grid used in the modelling has been selected to capture significant variations in concentrations. Thus, the method of interpolation will not have a significant influence on the contours. Tests have been done in the past to check this, but are not repeated for every modelling study carried out.

4.1.53(2) Can we see annual mean NOx concentration maps with contours at 1 µg m-3 intervals for all the modelled scenarios?

See attached plans showing 1µg/m3 contour intervals from 20 to 35µg/m3 for the 25 and 35mppa cases. Contours above 35 and below 20µg/m3 omitted for simplicity (in electronic copies of this rebuttal these plans are separate files “Appendix A 25mppa NOx 2003 Met 1µg intervals.pdf” and “Appendix A 35mppa NOx 2003 Met 1µg intervals.pdf”).

26


© Entec UK Limited

14899 May 2007

4.1.53(3) Was a NOx source apportionment carried out for sites within the forest? We would like to see the modelled airport NOx contribution for receptors along the forest’s northern edge and also at locations within the forest further away from the airport.

Dr Haycock to provide OS co-ordinates of the receptors for which concentrations are required, as at 21st May no reply had been received.

4.1.53(4) The receptor taken to be representative of Hatfield Forest in the report “Estimates of nitrogen deposition in Hatfield Forest and Eastend Wood for the Stansted G1 Application (ED47179/N/005)” appears to be north of the forest boundary. Why therefore was this receptor selected to be representative of the forest?

The location was the nearest 100m grid point to the extreme corner of forest boundary, albeit on the airport side of the boundary. This location would produce a higher deposition rate than at other locations within the forest. The airport NOx contribution decreases with increasing distance from the airport.

The following questions are focussed on the assumptions made within the air quality model, in respect of NOx (NO+NO2). What is the cumulative error of these assumptions on the overall predictions of mean annual NOx concentrations.

Note these paragraph references refer to the model test report.

2.6.2 It is assumed that vehicles issued with Temporary airside passes have a negligible effect on NOx emissions. How many of these passes are issued annually? Has any information been gathered about how long these vehicles usually operate airside?

Around 2000 temporary passes were issued in 2003, each valid for a maximum of 3 days, giving a maximum of 6000 vehicle-days. Total number of vehicle-days for permanent passes was around half a million, so temporary passes account for around 1% of the airside vehicle days. There is no reason to expect that temporary-pass vehicles have disproportionately higher fuel use per day on average than permanent-pass vehicles. In addition, only fuel brought from off-airport by temporary passes will be omitted from the emissions estimates: if any temporary-pass vehicle takes fuel from the airport station, its contribution is already accounted for.

In the 35mppa case these vehicles would contribute emissions of up to 0.4te/y of NOx which at High House would increase the annual mean concentrations by up to 0.02µg/m3. This is clearly a very small contribution.

2.6.5 What were the changes in the airside fleet between the 2003 activity period and 2003-4 monitoring period?

Little or no change in airside fleet would be expected between 2003 activity period and the 2003-04 monitoring period used for the model test.

2.6.9 Question withdrawn.

2.6.12 What are the differences in mean Euro standards for different size ranges?

The differences in the emission factors of vehicles complying with the standards are set out in Table 2.9 of the methodology report (CD/190).

27


© Entec UK Limited

14899 May 2007

2.6.21 “The omission of cold start emissions is expected to be minor compared to hot exhaust emissions”. What data have been collected to verify this assumption?

If every Light Duty Vehicle (cars, vans and other light duty vehicles) - the only type for which cold-start emission penalties are known - operating airside had five cold starts every day of the year (surely an overestimate) the total cold start emissions would be less than 1% of the airside emissions already calculated.

2.8.5 Taxi and bus queuing/idling emissions are ignored. How many vehicles are involved? If this is not known, how can their emissions be said to be insignificant?

Past experience of quantifying these contributions at other large airports (Heathrow and Gatwick) has shown them to be small sources, even if idling times of order tens of minutes per vehicle are assumed. Generally, vehicles are encouraged to switch off their engines when waiting. At Stansted with around half a million taxi transactions in the year, each idling for 10 minutes would produce (for large diesel cars) an emission of <1 tonne NOx per year, which is (much) less than 1% of the total ground-level NOx emissions on the airport. The contribution from buses (even if they did idle for this length of time, which is unlikely) would be much less.

4.5.19 Why were different data sources used for non-airport traffic volumes in the 15+ and current assessments?

We used the best source of background traffic data available at the time. In the 15+ assessment, 2002 was a forecast year (for which measured data were obviously not available) but when we did the 2003/4 assessment there were measured traffic data available for 2002 (which are incorporated into the NAEI data set used for the assessment).

4.1.17 Known entrainment effects were not included in the modelling on the basis that they would not contribute significantly to annual mean NOx concentrations although they can be significant on a daily basis. How has this effect been incorporated into any metrics used to calculate short-term concentrations from the modelled annual mean?

The MMU4 work indicates that the contribution would not be major on a daily basis. This entrainment effect (as previously noted 4.1.17 page 23) was not taken into account when calculating short period concentrations (hourly NO2; 24-hour PM10). The increment change due to this effect would be small and limited to areas close to the flight path. The maximum ground-level impact of material transported downwards in vortices will tend to occur in different locations from the impact of ground-level sources on the airport, so is unlikely to contribute to exceedances of standards.

4.1.25 “More accuracy may be needed (in seasonal and diurnal emission profiles) if model results are needed for shorter averaging times”. How has this been accounted for when calculating shorter-term metrics from annual means?

The model was not used directly to obtain short-period concentrations. In using the empirical relationships it was assumed that the temporal variation of sources on and around Stansted does not differ sufficiently from that of the sources influencing the monitoring data (on which the relationships are based) to invalidate their use (and see the answer to E3).

28


© Entec UK Limited

14899 May 2007

Stansted Airport Generation 1 Inquiry Rebuttal …stopstanstedexpansion.com/documents/BAA_Air_Quality...BAA/4/D (Case Reference No: 2032278) Stansted Airport Generation 1 Inquiry Rebuttal

Documents