H1 Annex E – Surface Water Discharges (complex)

Published by:

Environment Agency Horizon House, Deanery Road Bristol BS1 5AH Tel: 0117 934 4000 Email: [email protected] www.environment-agency.gov.uk

© Environment Agency

All rights reserved. This document may be reproduced with prior permission of the Environment Agency.

Further copies of this report are available from our publications catalogue: http://publications.environment-agency.gov.uk or our National Customer Contact Centre: Tel: 03708 506506

GEHO0410BSIK-E-E v2.1

Annex E Surface Water Discharges (complex)

This guide gives advice on assessing impacts of complex surface water discharges arising from the operation of sewage treatment works or from discharges to groundwater when applying for a bespoke permit under the Environmental Permitting Regulations. It is part of the Environment Agency’s H1 Environmental Risk Assessment Framework and should be read in conjunction with the H1 Environmental Risk Assessment – Overview guide to understand who needs to use it and how it fits in to the framework.

Content

H1 Annex E Surface Water Discharges (complex) v 2.1 December 2011 1

Summary of Changes.................................................................................................................... 2

About this guidance ..............................................................................................................3 Who does the assessment ...................................................................................................3 Overview...............................................................................................................................4 Conservation sites ................................................................................................................5 Continuous discharges of trade or sewage effluent..............................................................7 Proposed new or increased discharges..............................................................................13 What you should do next ....................................................................................................18

Appendix A Calculation of River Needs Permits .................................................................... 19 Appendix B Calculating Permit Limits for Priority or Hazardous Substances ................... 38 Appendix C Evidence to show storm overflow discharges are unsatisfactory.................. 58

.

Summary of changes


Summary of Changes

Below is a summary of changes made to this Annex since the launch in April 2010. Annex version

Date Change Template version

Issue 2.1 December 2011

References on “no deterioration” reflecting changes to our regulatory approach in response to the Water Framework Directive 2000/60/EC. Changes to section on Priority and Hazardous Substances to take account of the EQS directive 2008/105/EC.

Improved methods for calculation of permit limits for discharges of Priority and Hazardous Substances.

Changes to Appendix A which deals with the use of mathematical models.

Responding to issues raised in the 2010 H1 Public consultation.

H1 April 2010

Introduction


About this guidance This guide on assessing discharges to surface water is a part of the Environment Agency’s H1 Environmental Risk Assessment1 framework. It is aimed primarily at water companies and other organisations with water quality modelling skills and technical knowledge and experience of effluent treatment. You should follow this guide if: • you are applying for a permit under the Environmental Permitting Regulations • you need to carry out a bespoke risk assessment • you are working through H1 Environmental Risk Assessment • you have been advised to assess discharges to surface waters from your activity You should not work through this guidance without first reading through H1 Environmental Risk Assessment – Overview to see how it fits in to the risk assessment process for permit applications and carrying out the necessary preliminary steps. You do not need to work through this guidance if your activity does not involve the discharge of substances to surface waters. Other guides on assessing surface water discharges The H1 framework has two annexes on assessing surface water discharges: • a basic method intended for most installations and waste operations • this more complex methods for continuous and intermittent discharges This guide is most likely to be appropriate for water companies and other organisations with water quality modelling skills and technical knowledge and experience of effluent treatment. Who does the assessment For most permits, including installations and waste sites, you will need to do the assessment or engage a consultant to do it for you. In any case you should contact us first for information and advice. For water discharge activity permits: For permits for continuous water discharge activities to inland rivers, you may do the assessment yourself, but if you prefer you can provide us with the details of the discharge you wish to make, and we may be able to do the assessment for you. Most of the information we will need to do this can be provided on the application form, but you may need to add other relevant details. You should contact us for information and advice first.

1 Environment Agency. (2010) H1 Environmental Risk Assessment – Overview. Environment Agency, Bristol. (Available on the Environment Agency website)


As a general rule, we cannot undertake assessments of discharges to estuarine or coastal waters on your behalf, but we can offer advice as to what needs to be done. If you do not have the skills to do this yourself, then you will need to engage a consultant to advise you and to undertake the assessment. For intermittent discharges of storm sewage we can advise what assessment work needs to be done, but where sewerage or environmental impact modelling is required, we are unable to undertake this work on your behalf. Overview This guide will help you work out whether your discharges present a significant risk to surface waters. It covers two kinds of discharge: • continuous, such as treated sewage or trade effluents, • intermittent, such as combined sewage overflows and storm tank discharges. We use two tests to decide if discharges to surface waters are acceptable. A discharge is usually acceptable if: 1. it does not cause deterioration in quality of the water body receiving the discharge. We will assess discharges using the ‘no deterioration’ test if:

• applying for a new permit for a new activity • applying to increase currently permitted discharges, and

2. the receiving water body meets its target quality standards. We always assess the effect of discharges against the uses objectives and target standards for the receiving water body. The ‘no deterioration’ principle and target standards are explained in more detail below. No deterioration When we review permits for existing discharges or issue permits for new ones our aim is to issue permits that prevent or minimise any deterioration in the quality of water bodies that could otherwise occur as a result of the discharge. We must also be sure the proposed discharges do not make it impossible to achieve any target standards not currently being met (such as the Water Framework Directive Status Objective). We refer to this as ‘no deterioration’ and our ideal is for no increase in the planned pollutant load discharged to the water body. Where this is not possible we will limit any within-class deterioration as far as possible. We must maintain the Water Framework Directive status of water bodies (as reported in the December 2009 River Basin Management Plans approved by Ministers). This may exceptionally require action beyond the requirement for no increase in the permitted load to the water body.


If the control measures you would need to achieve no deterioration are not practical or not cost-effective, we may either refuse the permit or require you to use technically feasible and cost effective measures. Target standards When we are seeking improvement in water quality, our objective is to make sure that the permits we issue meet the uses, water quality objectives, environmental quality standards and design standards applicable to the receiving water. These include the Water Framework Status Objectives. We apply the water quality standards arising from the Directives listed below: • The applicable Water Framework Directive standards for the receiving water body and

water bodies affected by the discharge http://archive.defra.gov.uk/environment/quality/water/wfd/documents/riverbasinguidance-Vol2.pdf

• Freshwater Fish Directive • Dangerous Substances Directive • Habitats Directive (see Habitats specific guidance) • Shellfish Waters Directive • Bathing Water Directive and revised Bathing Water Directive • Nitrates Directive • Urban Waste Water Treatment Directive • Urban Waste Water Treatment Regulations DETR Guidance (1997) Determinands used by the Drinking Water Directive and other determinands specified by water companies to protect water supplies in Drinking Water Protected Areas, must not deteriorate. Conservation sites You should identify all possible receptors and include them in your assessment, this includes nature conservation sites, heritage and landscape and European species that could be affected by your activity. You must assess risks to any conservation site or species: • within the boundary of your activity • outside the boundary of your activity where you could directly affect a designated

feature • or indirectly affect the conservation site's environmental quality Nature conservation sites can include:- • Special Area of Conservation (SAC) • Special Protection Area (SPA) • Ramsar • Site of Special Scientific Interest (SSSI) • National Nature Reserve (NNR)


• Local Nature Reserve (LNR) • Ancient Woodland • Local Wildlife Site. You should contact the Environment Agency or conservation agencies to discuss the scope of risk assessments for such sites.

Annex E – Surface water discharges (complex)


Annex E – Surface water discharges (complex) Continuous discharges of trade or sewage effluent To determine if continuous discharges are likely to be acceptable: 1. identify the uses objectives and target standards for the receiving water body of your

discharge. 2. assess if receiving water currently meets the reported and target standards 3. calculate allowable discharge limits 4. assess the change in size of the mixing zone 5. decide if it is feasible to meet these limits 6. check statutory requirements on emission limits 7. check non-statutory requirements on emission limits 8. confirm final discharge and controls These steps are discussed below.

Identify reported and target standards (Reported standards are those reported in the December 2009 River Basin Management Plans and approved by Ministers). To identify reported and target standards for the water body receiving your discharge and any other water bodies affected by your discharge: • identify relevant reported and target standards for each discharged substance or

determinand. • take the most stringent standard if more than one is available for a discharged

substance or determinand. You must contact the Environment Agency to obtain or confirm the correct reported and target standards that apply. Make sure that you also check effluent temperatures and Ph against appropriate standards and use the correct statistical basis for comparisons. You should provide a summary of this comparison and explain whether you think your releases are acceptable. Where your application relates to a significant increase in a discharge you should contact us in good time. Ideally you should contact us at least 12 months before a permit decision is needed for informing decisions on growth. This is because additional water quality data may need to be collected to check the reported Water Framework Directive Status for the water body.


Assess if receiving water currently meets its reported and target standards We can give you the current compliance against reported and target water quality standards. Calculate the allowable discharge limit You may be applying: 1. to improve your existing discharges to meet currently failed target environmental standards and/or 2. for new or increased discharges or increased discharges that meet the requirements of ‘ no deterioration’. The procedure to calculate permit limits depends on whether or not the receiving water meets the reported standards and the target standards for each identified substance or determinand. For most calculations of permit limits use the modelling tools specified in Appendix A “Calculation of River Needs Permits” (RNC). Procedure to calculate permit limits where the receiving water meets the reported or target standards: You need to work out the allowable discharge for each substance or determinand to achieve no deterioration. For calculation of permits that are liable to contain (see below) any of the substances in the table below, use the methodology set out in Appendix B Calculation of Dangerous Substances Permit Limits.

• Priority Substances identified in the Water Framework Directive, or • Hazardous substances under the Dangerous Substances Directive and its

daughter Directives, or • List I and List II substances, or • Specific Pollutants, or • any other substance that the Environment Agency has determined as hazardous.

Substance Where to find a list of these substances Priority Substances identified in the Water Framework Directive

The River Basin Districts Typology, Standards and Groundwater threshold values (Water Framework Directive)(England and Wales) Directions 2010

Specific Pollutants The River Basin Districts Typology, Standards and Groundwater threshold values (Water Framework Directive)(England and Wales) Directions 2010

any other substance that the Environment Agency has determined as hazardous

Non-Statutory Environmental Quality Standards – Table A2 of H1 Annex (d)


Liable to contain A discharge is described as liable to contain a substance if any of the following apply. Test Questions Substance is permitted or otherwise allowed to be discharged into a sewerage catchment upstream of the discharge.

• If there are any consented Trade Effluents discharging to sewerage system, do any have limits for Priority or Hazardous Substances (PS)?

Substance is known to be added as a result of activities within the sewerage catchment upstream of the discharge.

• Do we have any information that PS are used in the catchment and could be present in the sewage in significant quantities?

Substance is known to be added at the Permit Holder’s site.

• Does the operator use raw materials or effluent treatment chemicals containing any PS ?

Substance detected by chemical analysis in the discharge; within the sewerage catchment; or in the treatment process.

• Do we have any PS analysis results in any of these locations that have detected PS? (NB use the Detection Test table below)

Substance detected by chemical analysis in the receiving water, where we are confident that it has come from the discharge

• Do we have analysis results in the downstream watercourse that have detected PS which have come from the discharge? (NB Use the Detection Test table below)

If you answer ‘Yes’ to any question, the discharge is liable to contain a Priority or Hazardous Substance (PS). List all the PS that the discharge is liable to contain for further assessment. Only use actual information about the discharge under consideration. Do not make assumptions based on other discharges. Then go to Appendix B Detection Test You must be at least 95% confident that the Substance is present in the discharge, at or above the Minimum Reporting Value (MRV), for at least 10 percent of the time for the discharge to be assessed as liable to contain the substance. MRVs are usually expressed as 10% of the EQS-AA. Values of EQS-AA and hence NRVs may be found in Appendix A of Annex D. For the number of sample results in the left hand column, you should have at least the number of results in the right hand column above the MRV to be confident that this test is met:


Detection Test Table Number of Samples of Discharge Quality

Minimum number of results at or above the Minimum Reporting Value

2 – 3 2 4 – 8 3

9 – 14 4 15 – 20 5 21 – 27 6 28 – 34 7 35 – 41 8 42 – 48 9 49 – 56 10 57 – 63 11 64 – 71 12 72 – 79 13 80 – 86 14 87 – 94 15

95 – 102 16 How to achieve “no deterioration” You need to determine appropriate limits for the discharge. These limits must be environmentally protective, and technically feasible, for a cost that is proportional to the benefits. Our ideal is for no increased polluting load to the water body. to meet this:

• for proposals to increase the volume of an existing discharge the concentration of each parameter must be reduced proportionately, unless the effect of your discharge is trivial.

• For new discharges you need to manage all your discharges so that the overall polluting load does not increase.

Where this is not feasible or cost effective we may allow a within class deterioration of up to 10% in the current water quality To meet this:

• Include existing discharges in the modelling assuming that all are discharging at the current discharge loads and maximum practical loads allowed by the current permits as appropriate for the needs of the modelling. Note that the maximum should be assessed realistically. Do not, for instance, assume that all discharges will simultaneously discharge at the maximum permitted load.


• For each determinand, you may allow the receiving water to deteriorate by up to 10 per cent in the mean and 90th percentile quality (or 95th or other relevant percentile quality where appropriate).

• For rivers, calculate the downstream target water quality, as the current water quality (mean and 90th percentile/95th or other relevant percentile), plus up to 10 per cent of the current quality.

• Check that your discharge does not cause deterioration beyond the target standard. If this does happen you need to contact us.

• Check that your discharge does not cause deterioration beyond the reported standard. If this does happen you need to contact us.

Un-ionised ammonia If the receiving water body has target standard for un-ionised ammonia, the ‘up to 10 per cent’ rule does not apply directly to un-ionised ammonia. Instead, you should:

• determine the receiving water target for the total ammonia. • check to ensure that meeting this target would not breach the required standard for

un-ionised ammonia. The Environment Agency can provide you with an Excel spreadsheet to help you with this calculation (contact us and request document 200_05_SD02 WQ Permitting guidance on freshwater fish directive compliance.)

If the receiving water body does not have a target standard for un-ionised ammonia, ensure the permit limits you derive limit the deterioration to not more than 10 per cent in the mean and 90th and other relevant percentile concentrations of determinands in the receiving water after mixing, unless you can show that there is no significant environmental change as a consequence. Assess change in size of mixing zone Where there is a significant change in the discharged load you must consider the effect on the size of the mixing zone. The mixing zone is defined as the area within which the target standard is exceeded. The change in the size of the mixing zone must be acceptable and not impact on the uses and objectives of the receiving waters.

Decide if it is feasible to meet these limits You must decide if it is feasible for your to meet your discharge limits by considering: if it is technically feasible to achieve the permit limit you have derived for no

deterioration? if the cost of achieving the permit limit you have derived for no deterioration is

proportionate to the benefit? If you think that the answer to either of the question is NO you must contact us. Otherwise, you should carry on with the assessment. Check statutory requirements on emission limits Some activities, including larger discharges of treated sewage effluent and certain types of industrial discharge, also have statutory limits on discharges. You should check that the limits you have calculated meet the statutory limits. These are specified within the Urban Wastewater Treatment Regulations and you should contact us for these limits.


Carry forward the more stringent limits to the next section. Confirm final discharge and controls You should summarise the outcome of the above assessment, including: • the substances and determinands you intend to discharge • the limit of deterioration or target standard these meet • what controls you have in place and how they are justified (e.g. by CBA or 10%

deterioration). You will need to present your working and calculations.

Intermittent sewage effluent discharges This guide covers the following types of intermittent discharges: • combined sewer overflows; • storm tank discharges. It does not cover rainfall-related surface water discharges, nor emergency overflows from pumping stations. The extent of the assessment you will need to carry out depends on whether you are improving an existing discharge or creating a new or increasing the discharge from an existing overflow. Existing discharges You will need evidence to show whether these are unsatisfactory. You may already have been informed by the Environment Agency if your discharge is classed as unsatisfactory. Alternatively there is guidance in Appendix C on how to do this investigation. You should contact the Environment Agency to see if it holds any data to help confirm the satisfactory/unsatisfactory status of the overflow. If your discharge is not unsatisfactory you can apply for a Permit on the basis of the existing asset. Criteria for unsatisfactory overflows are: • causes significant visual or aesthetic impact due to solids, fungus; • causes or makes a significant contribution to a deterioration in river chemical or biological

class; • causes or makes a significant contribution to a failure to comply with Bathing Water

Quality Standards for identified bathing waters; • operates in dry weather conditions; • operates in breach of Permit conditions provided that they are still appropriate; and/or • causes a breach of water quality standards (EQS) and other EC Directives. In general unsatisfactory discharges will need to be improved to meet the relevant target standards. You will need to contact the Agency to confirm the appropriate standards.


These may be a combination of water quality standards, emission standards, and sewerage design standards. The Urban Pollution Management Manual 2nd edition published in October 1998 (FWR FR/CL 0009) provides methods for assessing your proposals against the required standards. Changes to existing discharges that are currently satisfactory should not cause any deterioration in the quality of the water body. The changes should also undo any historical deterioration that has occurred since the structure was originally designed and built. Proposed new or increased discharges These are generally applied for either: • to reduce flooding • allow an existing emergency only overflow to operate in storms • to replace existing overflows as part of an overall water quality improvement scheme. • to limit flows to full treatment at a sewage treatment works. If the proposal is to reduce flooding or allow an existing emergency overflow to operate in storm then contact the Environment Agency as we do not usually permit these. We prefer excess surface water and infiltration to be kept out of foul or combined sewers. If part of an overall scheme to improve intermittent discharges then any localised deterioration may be balanced against the improvements – you will need to consider any affects on amenity, or recreation. If a new or increased overflow is required to help improve final effluent quality by limiting the flows to treatment we will expect the overflow to meet the relevant sewerage design standards. Additionally we will expect water quality standards to be met and there to be no deterioration in the quality of the water body in wet weather. To do this you will need to determine the net effect of improving the fully treated effluent during storms whilst increasing storm sewage spills.


The following table summarises these requirements:

Purpose of overflow

Meet Design

Standards

Meet Water Quality

Standards

Demonstrate No

Deterioration Normally Refuse

Existing overflow satisfactory

Yes

Existing unsatisfactory

Yes Yes

New/upsized storm tanks

Yes Yes Yes

Limit flow to treatment As part of improvement scheme

Yes Yes Yes

Reduce flooding Allow pumping station emergency overflow to discharge in storm

Yes Yes Yes Yes

Action

See below Contact EA And see below

See below Contact EA

Design standards Definitions For design purposes the following definitions apply: P = Population served G = Water consumption per head per day I = Infiltration (maximum) E = Trade Effluent flow to sewer as applicable Flow to full treatment The flow passed forward to full treatment is controlled by the setting of the final overflow on the sewerage. This setting known as the flow to full treatment (FFT) and should exceed the maximum flow rate arriving at the works in dry weather conditions. It should also be sufficient to allow any storm storage within the sewer system and at the sewage treatment works to be emptied within a reasonable time after rainfall. It should be designed so that storm storage can be emptied soon enough to be available for subsequent events. You will need more storage to meet water quality standards if the FFT is set too low. This is because additional storage will be needed to retain closely spaced rainfall events. This will be more costly overall.


Normally flows up to 3DWF (defined below) arriving at the sewage treatment works should be passed to full treatment. This should allow storm tanks at the sewage works to be emptied as soon as practicable which is a Permit requirement. 3DWF is defined as 3PG +1 +3E The factors in this formula should be based on the operator’s design horizon and should be validated against any available measured flow data. Storm tanks These standards apply to flows in excess of the FFT arriving at sewage discharging inland waters and estuaries. These design standards are normally considered as minimum standards. Higher or lower standards may be appropriate depending on receiving water quality standards and any no deterioration requirements. Storm tanks: • Should only be filled whenever incoming flows exceed the (FFT) • Should be properly designed to settle suspended solids • Should have a storage capacity of either 68 litres/head of population served or 2 hours

storage for the maximum flow received. • Should not discharge a significant quantity of solid matter having a size greater than

6mm in any one dimension. This requires some form of solids separation having a performance equivalent to a 6 mm 2 dimensional aperture screen See UPM Manual 2nd edition for details.

• Their contents should be passed to full treatment as soon as practicable after it has stopped filling. This usually requires an automated system to return stored flows whenever the flow arriving at the sewage works is less than the FFT.

Combined Sewer Overflows (CSOs) To protect amenity values in the vicinity of the discharge it should not contain a significant quantity of solid matter having a size greater than 6mm in more than one dimension. This requires some form of solids separation device that has a performance equivalent to a 6 mm 2 dimensional aperture screen, see UPM Manual 2nd edition for details. The Agency may accept a reduced level of protection in situations where the amenity value in the vicinity of the discharge is low. Discharges of storm sewage from combined sewer overflows should only comprise of flows over and above the design pass forward flow (PFF). The normal minimum requirement for the pass forward flow is given by Formula A which is defined as follows: Formula A (litres/day) = (PG+I+E) + 1360 P + 2E Alternatively, the operator may use sewer modelling to demonstrate that the frequency and volume of spill from the proposed overflow will not cause significant deterioration.


Additional requirements for CSO discharges to Bathing Waters: Where a discharge impacts on a EU designated bathing water these overflow should be designed so as not to spill no more than 3 times per bathing season on average. The bathing season is May to September inclusive. As a minimum the most recent 10 year rainfall record representative of the rain falling on the catchment should be used for this analysis. Guidance on this can be found in the UPM Manual 2nd edition and from the Environment Agency. Where more than one overflow impacts on the same part of the bathing water the total spills from those overflows should not exceed 3 per bathing season on average. Guidance on how to aggregate discharges is available in from the Environment Agency. For the purpose of this guidance a single spill is defined as any discharge or series of discharges occurring within 12 hours. If the discharge/s lasts between 12 and 36 hours it counts as 2 spills. Any discharge/s extending beyond 36 hours count as an extra spill for each additional 24 hours period. Individual spills counted in this way are then discounted if the volume discharged to the bathing water, as totalled from all overflows, is less than a particular significance threshold. You will need to contact the Agency to confirm the threshold which may be site specific. Discharges greater than 50 m3 will usually be significant. Additional CSO requirements for discharges to Shellfish Waters: Where a discharge impacts on a EU designated Shellfish Water these overflow should be designed so as to spill no more than 10 times per year on average. As a minimum the most recent 10 year rainfall record representative of the rain falling on the catchment should be used for this analysis. Guidance on this can be found in the UPM Manual 2nd edition. Further guidance on this approach is also contained in the Agency document: Consenting Discharges to achieve the requirements of the Shellfish Waters Directive (microbial quality) 169_01. The counting of spills and discounting of minor discharge volumes is as described above for bathing waters. Water quality standards You may need to check whether existing discharges comply with the relevant water quality standards or you may need to design schemes to meet the appropriate water quality standards. These include standards arising from EU Directives as well as 99 percentile standards and fundamental intermittent standards which are used for design purposes. You should contact the Environment Agency to confirm the standards relevant to the receiving water. The Environment Agency may also have flow and quality data to generate boundary conditions for use in modelling.


The UPM Manual 2nd edition describes methods for demonstrating compliance and designing improvements. The proposed method, models, and data, should be agreed with the Agency before work commences. All models should be confirmed as fit for purpose by independent review and audit. Discharges to Bathing Waters For discharges to EU designated bathing waters it may be possible to design improvements to meet the microbial standards of the EU bathing water directive as an alternative to the spill frequency standard described above. You should contact the Agency to confirm whether this is an option and if so what the appropriate standards are. Discharges to Shellfish Waters For discharges to EU designated shellfish waters it may be possible to design improvements to meet the microbial standards of the EU Shellfish Waters Directive as an alternative to the spill frequency standard described above. You should contact the Agency to confirm whether this is an option and if so what the appropriate standards are. Demonstrate ‘No Deterioration’ The Environment Agency’s objective when permitting new or increased discharges is for no deterioration in the wet weather quality of the receiving water. In general for rivers and estuaries this means no deterioration in the quality of the water at any percentile, including the 99th percentile and higher, and no increase in the number of exceedences of the `fundamental intermittent standards` set out in the UPM Manual 2nd edition. Relocating or rationalising and reducing storm discharges may cause localised deteriorations which may be outweighed by improvements elsewhere. You should consult with the Environment Agency to see whether this is likely to be acceptable. Creating a new or increased storm overflow at the inlet to a sewage treatment works will only be considered where the impact of the additional storm discharge is outweighed by the improvement in the wet weather quality of the fully treated effluent. This can be demonstrated using modelling to carry out a full impact assessment. Alternatively a simplified approach involving only sewerage modelling is described below: The simplified approach is set out below, other approaches acceptable: • Predict a time series of the existing final effluent load by running the existing system

model for a time series of rainfall events. • Predict a time series of the combined load from the storm tank and the improved final

effluent by running the rainfall time series on the proposed system model. • Compare the two load time series at various percentiles, including ones greater than

the 99th percentile, to make sure there is no deterioration in the discharge load at any percentile.


What you should do next If you are happy that you have completed your assessment of surface water discharges correctly, you should: • complete any other risk assessments for your activity, as set out in Step 2 of H1

Environmental Risk Assessment – Overview • if you have assessed all the risks from your activity, continue with Step 3 of H1

Environmental Risk Assessment – Overview. You should also document your assessment of releases to surface waters.

Appendix A Calculation of River Needs Permits


Appendix A Calculation of River Needs Permits Introduction

This appendix is about calculating a particular kind of numerical limit that is placed in permits for discharges to rivers. These permit limits are standards that need to be achieved by the discharge in order to meet specified numerical values for water quality standards2 in the river. We call these types of permit limits – River Needs Permits.

Such limits will be required where the discharge load is appreciable in the context of the standard that must be achieved in the river3, in terms of the dilution provided by the river.

Some types of limits in permits are not set to meet particular numerical standards for water quality in the receiving water. These types of limits may take be limits imposed on particular types of discharge or industry, for a variety of reasons and risks linked to legislation. This appendix does not apply to these types of limits, though, in terms of topics such as the correct assessment of compliance, there are common features.

This appendix is also a guide for the Environment Agency’s software for calculating such permit limits. This software is called RQP (River Quality Planning). It is available from the Environment Agency on a CD or memory stick.

This appendix is based on the Environment Agency’s guidance to its staff. The full version of the guidance can be obtained from the Agency.

In particular, this appendix covers the permit limits that are needed to achieve the water quality standards defined for Directives issued by the European Commission.

Some of the standards which define the High, Good, Moderate and Poor classes for the Water Framework Directive are given, by way of illustration, in Table 1.1 to 1.4 for two of the common determinands. You will probably need the advice of the Agency on the particular standards, and which determinands, apply for you.

Table 1.1: Standards for ammonia

Total Ammonia (mg/l) (annual 90-percentile)

Type of river High Good Moderate Poor

Upland and low alkalinity

0.2 0.3 0.75 1.1

Lowland and high alkalinity

0.3 0.6 1.1 2.5

2 Such standards may also include requirements to meet “standards” based on limiting deterioration. 3 The term “river” is used throughout thought the principles apply to all discharges whose permit limits are dictated by dilution


Table 1.2: Typology for nutrient conditions for rivers

Annual mean alkalinity (as mg/l calcium carbonate)

Altitude

Less than 50 More than 50

Under 80 metres Type 1 Type 3

Over 80 metres Type 2 Type 4

Table 1.3: Standards for phosphorus in rivers Soluble Reactive Phosphorus (μg/l) (annual mean)

Type (Table 1.3) High Good Moderate Poor 1 30 50 150 500

2 20 40 150 500

3 and 4 50 120 250 1000

The next part of this appendix deals with the basic calculations. Most users will seldom need to step beyond these.

Our broader guidance (see para 5 on page 19) covers the calculation of the permit limits needed to meet river standards for un-ionised Ammonia. It also includes intermittent discharges, seasonal limits, catchment models like SIMCAT, and other details. The Basis of the calculations

The mixing of a discharge with a river is described by the Mass Balance Equation [1]:

fFfcFCT

++

=

where: • F is the river flow upstream of the discharge • C is the concentration of pollutant in the river upstream of the discharge • f is the flow of the discharge • c is the concentration of pollutant in the discharge • T is the concentration of pollutant downstream of the discharge.

The most commonly used quality standards for rivers are the annual 90-percentiles for the Biochemical Oxygen Demand (BOD) and Total Ammonia and the annual mean standard for Phosphate.

The 90-percentiles are standards that will be achieved with the required degree of reliability so long as the concentration specified as the 90-percentile limit is exceeded for no more than 10% of the time.


Similarly, the discharge limits for BOD and Total Ammonia that come out of the calculations will always be expressed as annual means or annual percentiles. There is no avoiding this in a correct calculation.

For BOD and Total Ammonia the discharge limits are set as annual 95-percentiles in the permits for sewage treatment works operated by the water companies. For phosphate they are usually set as an annual mean.

At the time of writing, the values of the annual 95-percentile limits we calculate for all other determinands in the permit are actually placed in the permit as if they were absolute or maximum values. (This may change in future).

But in these we seek to interpret the absolute limits as the original annual 95-percentiles when, for example, we look at the question of the performance of a discharge in meeting its permit and protecting the river.

It is a mathematical fact that a single application of Equation [1] cannot be used to calculate the permit limits needed to meet river targets4. We must use methods that produce the statistical distribution of T by combining the statistical distributions of F,C, f and c.

We do this by Monte-Carlo Simulation. In this, a value for each of the variables F, C, f and c is plucked randomly from the full range of possible values.

A value for T is calculated from each set of plucked values of F, C, f and c using the Mass Balance Equation. The sequence of selection and mass balance is repeated until enough values of T have been calculated to define its full statistical distribution. Each value of T (or each value of F,C, f or c) is called a shot. Routine calculations are set up to use 1000 shots for T.

It is usually sufficient to assume that F, C, f and c follow the log-normal distribution, but any other distributions can be used.

You will be aware that the values of F, C, f and c are not random in practice but may be related to one another. For example, after rain, the values of the river flow, F, and the discharge flow, f, will rise together. We say that F and f are correlated.

You can introduce all the statistical correlations between F, C, f and c by entering values for correlation coefficients. RQP contains a default correlation coefficient of 0.6 between river and discharge flow. The default values are zero for all the other correlations.

In most cases you will not need to change these defaults. We discuss later when you might want to.

4 Other methods produce permit limits that are not those needed to meet the river quality standard. Usually they are too strict.


DOING THE CALCULATIONS

When you start RQP you get the following screen. It lists the various types of calculations you can do.


Click MONTE CARLO SIMULATION to start. You need to enter a value in each blue box.

The calculation is then done and the results displayed on the screen. The method may be used “forwards” (to calculate the river quality downstream of some particular discharge quality). Click Calculate effect of input discharge quality for this.

It can also be use “backwards” (to calculate the discharge quality needed to achieve a target quality in the river). Click CALCULATE REQUIRED DISCHARGE QUALITY for this.

An example of the output for the forward calculation is shown below. A 95-percentle discharge quality of 27.8 produces an annual mean in the river of 4.9.


The results can be printed by clicking Print on the File menu. You can transfer them to your own reports using Save results to spreadsheet file.


Monte Carlo Simulation can test the effects of assumptions by clicking Further Data.

This produces the screen shown below.

This can be used as a check on the sensitivity to the assumption that the river flow is log-normal. Change the shift flow to say, half the 95-percentile, to check this. (Note that, as shown on the first screen produced by RQO, that there are options to use any distributions for F, f, C and c).

Also, you can see that this screen can be used to calculate the effect on river quality of abstractions that reduce the flow in rivers upstream of discharges. It can also evaluate options that might improve river quality by adding to river flow upstream of a discharge.

There is also a “sensitivity analysis” that produces a report on the effects of lots of changes in the data.


DATA

You will have seen above that you need data that characterise the statistical distributions of F, C, f and c. In most cases, the data can be presumed to be log-normal. This means that two summary statistics will define the complete distribution. Any two statistics may be used, so RQP uses those readily available. These are:

River flow: mean and 5-percentile Upstream river quality: mean and standard deviation Discharge flow: mean and standard deviation Discharge quality: mean and standard deviation


Target River Quality

The river target is usually an annual 90-percentile or an annual mean although other percentiles could turn up for various pollutants. Water quality standards that are expressed as things like absolute limits, or Maximum Allowed Concentrations, are taken to be annual 95-percentiles5. General Background Most users just obtain or compute the summary statistics and get on with the calculations. This would make use of the built-in assumptions about distributions and correlation.

In most cases this produces good results. The results, and the decisions on permit limits which are based on them, are not significantly affected by the built-in assumptions.

You might like to use packages like EXCEL to inspect the raw data in the form of graphs. This will show up peculiarities, like outliers, which might have produced summary statistics which were misleading.

You might want to use things like EXCEL to display the data as histograms, and to test which distributions provide the best fit to the data. This will show the cases where it might be risky, or too cautious, to assume a particular distribution. This advice is can be important for the less common pollutants.

You will probably be aware of the statistical errors associated with estimates of means and percentiles of river and discharge quality. These are nearly always more important than doubts about distributions and correlation.

You can calculate confidence limits on the estimates of percentiles by clicking on CONFIDENCE LIMITS ON ESTIMATES OF PERCENTILES on the RQP screen. There are other devices bundled up with RQP that allow you to calculate confidence limits for the annual mean, though it is simple to include such method in the process by which you calculated your annual means.

Where you have to do calculations despite gaps in knowledge you should make neutral, not pessimistic, assumptions (for example by sticking to the default options that are already set up in RQP). But you should check the sensitivity of your results to the uncertainties in data. Often you will find that these tests do not affect the results of the calculation too much, nor the decision on the permit limit. Cases where this is not the case are discussed below (see ‘Results which are sensitive to data’ on page 34)..

The main reason for avoiding pessimistic assumptions is that they lead to permit limits that are too strict solely. This outcome is unfortunate if the reasons are based on failure to resource the effort needed to collect data. It is sometimes too tempting to make pessimistic assumptions for several aspects of the data, for each of the four variables, and the river quality target. Such a combination is statistically unlikely in the Real World, and will result in proposals for permit limits which are much stricter than those actually needed to meet the river target.

5 This is because such absolute limits, coupled with the usual sampling rate, are not absolute limits, but mathematically closer to 95-percentiles than 99.999-percentiles.


A key pointing all this is that the derivation of the annual 90-percentile standard embodies both the impact of the pollutant and the acceptability (or risk) of the impact. It covers the risk of damage and has taken account of variation in water quality in the context of this risk.

If you are unhappy to assume the log-normal distribution you can try others. Experience suggests that you will not need to do this very often. The button on the first screen and labelled “MONTE CARLO SIMULATION FOR NON-PARAMETRIC DISTRIBUTIONS” allows you to use any distribution for river flow provided you spend a half-hour sorting out the data for this. It also allows any distribution for river quality, discharge flow or discharge quality.

Some discharges enter a stream close to its confluence with another watercourse. Here, the calculation may need to take account of the river quality targets of both rivers. This is mentioned below (see page 37 ‘Discharges to Tributaries’).

Sometimes you will deal with a set of discharges to a river and decisions on one discharge have a big effect on what should be done at the others. You should be able to deal with this in a systematic series of calculations starting at the upstream discharges and working downstream. You may need to allow for natural purification as the river flows downstream. For the most complex catchments it might save time if you use a Catchment Model like SIMCAT.

River Flow

For most discharges, estimates of upstream river flow will be available from past calculations. You may need to ask the Environment Agency for information on river flow.

Usually, two summary statistics are required to characterise the river flow distribution. The mean and the 5-percentile flow are easiest to use. With the assumption of log-normality, these statistics define the full Flow Duration Curve.

The statistics should be estimates for the past few years. Flows should have been adjusted for the net effect of current (or projected) discharges and abstractions.

In river quality planning, the low flow, like the 5-percentile low flow, is required, with the mean, solely to characterise the total distribution of flows6. Results are sensitive to gross errors in the flow data but are not so dramatically affected by error in the low flow value itself.

This means that the flow data need not be as accurate as required for, say, deciding an abstraction license. Where gauged data are lacking, estimates based on a knowledge of catchment area and type will usually be sufficient.

Small streams should not be given a zero low flow just because they are of trivial importance to water resources. If a better estimate of the low flow is, say, 1 or 2% of the mean then this value should be specified even though this might be viewed as zero as a water resource.

6 It is sometimes suggested that the 5-percentile low flow is used as the sole measure of dilution in the calculation of the permit limit. Such methods are wrong.


If you intend to use non-parametric distributions for river flow you will need to obtain an estimate of the full Flow Duration Curve (as a graph or a table of values). You can then take as many points from this curve as you need to characterise its shape. The points should be equidistant in respect of the probability that they occur.


Upstream River Quality

If you check the sensitivity of your results to the data you will find often that the data on the upstream river quality are the most important source of uncertainty.

It is sobering to calculate the statistical uncertainty in estimates of percentiles. Take Ammonia for example. An estimate of a 90-percentile based on 30 samples can easily have a 90% confidence interval which extends from -50% to +100% of the value estimated.

This uncertainty is caused mainly by the Laws of Chance in sampling – statistical sampling error. There may also be additional errors from sampling that is not representative and in chemical analysis.

The mean and standard deviation are required for river quality just upstream of the point of discharge. To avoid the effects of unusual years, and to minimise the effects of statistical sampling error on small sets of data, it is usually best to use results obtained, say, over 3-4 complete years.

Unless, of course, there has been a step change in river quality during this time. This might result from some improvement in the quality of the upstream river. In these cases it may be better to place the upstream river in the middle of its presumed class.

Where the nearest upstream monitoring point is some distance upstream of the discharge, it can be important to allow for the effects of natural purification from the monitoring point to the point in the river just upstream of the discharge.

To calculate this effect you may need to find two sampling points on the river (or a neighbouring river) between which there are no significant inputs of pollution or dilution. You then need the mean quality at both these points. You might use these data to work out a rate constant which you can then apply to work out the effect of natural purification on other stretches of river or on other rivers.

Where no upstream quality data are available, or when there are only a few samples, we can use data from neighbouring or similar rivers.

Alternatively, it is usually better to assume, at least for the BOD, that the upstream river lies in the middle or in the best quarter of its (presumed) Class. Faced with the sampling errors noted above in para 2, we may feel safer with estimates of upstream quality which are based on the presumed Class.

In the absence of any data on Ammonia which can be assumed to apply for the upstream river, we might use values which are near zero (or as recorded for clean streams).

We would not normally require a discharge to have tight standards just because pollution upstream of the discharge provides poor quality river water for dilution. We would expect that the upstream pollution would be improved so that the stretch of river achieves its target. If we need to consider the downstream discharge in advance of this happening, we might assume the upstream river lies in the middle of its target class, or easily complies with the water quality standard.

Also, we might use a Catchment Model to look at permit limits in one go for all the discharges in the catchment, along with other sources of pollution and damage. Discharge Flow. You need the annual mean daily flow and the standard deviation. For discharges which are gauged you should avoid data for part-years because of the seasonal variation.


For quickness, we often estimate for sewage works the mean daily flow as 1.25 times the stated Dry Weather Flow. The standard deviation is seldom a critical factor, and can often be taken as one third of the mean. Both of these factors will be larger for works which are overloaded.

It is wrong to assume the standard deviation is zero, just because you have no information.

When a works is to be extended, you will have to use estimates of the future flow. Here, it is usually adequate to base the estimates on the expected Dry Weather Flow of the new works, or by scaling the current flows.

In water quality planning it is usual to express the discharge load in terms of its flow and concentration and how these are correlated from day to day. Within the Monte Carlo Simulation this gives the appropriate statistical distribution of discharge load. It may be that for some types of discharge that the discharge flow is trivial compared with the river flow and that the “load” is dominated by the variation in concentration. In such a case it need not matter if the discharge flow is entered as a constant of 1.0 and the “load” entered as “concentration”. Seasonal Discharges

Some discharges do not operate all the time but only occasionally, or only at certain seasons, or on particular days of the week. These discharges should be handled using the Non-Parametric form of Monte Carlo Simulation that is listed on the first screen of RQP. In such a distribution the correct proportion of values will be entered as zero. Details are in the full version of the Environment Agency’s procedures. Discharge Quality

In the calculation of the permit limit (the backwards calculation), you may be surprised that the calculation asks you for the discharge quality. After all, you want to calculate the discharge quality needed to meet the river target.

The values you enter for the backwards calculations are used only as a starting point. It is important only that the ratio of mean to standard deviation is sensible. Outside this constraint, the actual values of the mean and standard deviation entered do not affect the results of the calculation of what is needed to achieve the water quality standard..

You can start with the mean and standard deviation for the current discharge quality. If the calculated permit limit turns out to be a lot tighter (or laxer) than the current quality you should repeat the calculation. This time start with the mean and standard deviation for a discharge quality which is typical of the type of works which will be needed to achieve the calculated permit limit. (We do this because the extent to which discharge quality varies about the mean does depend on the type of treatment). Seasonal Permit Limits

For some continuous discharges you may want to have water quality standards on the discharge which differ according to the time of year. The method for calculating these is in the full version of the Environment Agency’s procedures. Generally the calculations show that such limits hold few benefits to dischargers, compared with all-year limits.


Precision Required of Data

Generally, the sensitivity of your results to data is far smaller than the statistical errors associated with sampling programmes. Exceptions are discussed in below.

Two general points are made here:

• is quicker to repeat a calculation to gauge whether the permit limit is sensitive to the data, than to seek out and process input data to a prejudged impression of the required accuracy;

• any sensitivity to data reflects real uncertainty about the decisions needed to

achieve the river target. It is not a consequence of the method of calculation.

DOING THE CALCULATIONS

It is straightforward to enter your data and calculate the permit limit. It is advisable do more than one calculation. They take so little extra time.

It is helpful to check the downstream river quality produced by common discharge standards like 20 mg/1 of BOD and 5 mg/1 of Total Ammonia.

Important tests are: • the change in the permit limit caused by a small change in the river target or

the upstream river quality • the change in the downstream river quality caused by a variation of the

discharge quality and flow.

When the data are poor and the calculations give a strict permit limit, you should also check the effect of:

• a change in the ratio of standard deviation to mean in the discharge quality • a more optimistic assumptions about the quality of the upstream river.

Similarly, you may want to check the effect of correlation, especially that between river flow and:

• discharge flow; or, • discharge quality (especially for Ammonia).

If your decision on the permit limit for an important discharge is sensitive to correlation then it is prudent to calculate the actual correlation coefficients that apply to your discharge. To do this you need to go back to the raw data you used to calculate the summary statistics. You then need to pair these by date and, for cases where you assume a log-normal distribution, do a linear regression on the logarithms of the paired data. The coefficient provided by this, commonly called r, is the one required by as the correlation coefficient in the Monte Carlo calculation.

Finally you may want to satisfy yourself that the calculated permit limit is insensitive to the assumption that river flows are log-normal. We mentioned this above on page 25 after the Monte Carlo screen.


It is useful to calculate the river quality produced downstream of the current discharge (the forwards calculation). This shows the effect of the discharge relative to other sources of pollution and helps flag cases where the upstream quality is already close to the water quality standard we want to achieve downstream.

You may be able to compare this estimate of river quality with an estimate based on actual measurements of river quality downstream of the discharge. This may assure you that the calculations are correct, but you should be prepared for differences if only because of the effect of statistical sampling error in the data for river (and discharge) quality (see page 30 para 2).

The calculated river quality may be a better estimate of the true summary statistics for river quality downstream of the discharge than that obtained from the analysis of samples of the river water itself. This may happen because of the precision added by using the samples of discharge quality to estimate river quality.

In some cases there may be significant differences between the results of the your calculation and the measured quality in the river. This may indicate errors in the data used in the calculation, or the action of sources and sinks of pollution (and the effects like incomplete mixing). These factors may need to be taken into account in plans to meet river targets.

It may also indicate more subtle effects like those caused by the fact that we take most samples between the hours of 10.00 and 15.00. Because of the time taken for a volume of river water to travel, day-time river monitoring may tend, for example, to record in the river the effects of the effluent load discharged at night. This called the effect of Sample Window.

INTERPRETING THE RESULTS

The following paragraphs describe some of the types of results that require special care in interpretation. Failure to Achieve the River Target

The quality which you entered for the river upstream of the discharge can sometimes be worse than the river target required downstream. The permit limit will then have to be similar, or better, than the quality of the upstream river. In some cases it may be impossible to meet the river target even if the discharge quality had a concentration of zero.

The interpretation of this result is that the river target cannot be met unless river quality is improved upstream of the discharge. However, this type of result can be produced if you:

• use a river target which is too strict • neglect to allow for improvements to upstream sources of pollution; or, • ignore natural purification.

In these cases the results point out where a sensible permit limit can be set only if you consider the combined effect, say, of several discharges


Results which are Sensitive to Data

This condition will occur in many calculations for example when the upstream quality is close to the river target, there is only a small capacity for the river to take more pollution and still remain within the target. This capacity may increase several fold if the data for the calculation are changed slightly.

You should look at the data for the quality of the river upstream of the discharge and check whether this is close to the target required downstream of the discharge.

The calculated permit limit can also be sensitive to data in cases where the discharge is so small that it has a trivial effect on river quality. What may happen is that the calculated permit limit, which may be a huge value, will change markedly following a slight change to the river target or upstream river quality.

Often the practical effect of this will be limited by the use of off-the-shelf minimum stan-dards for permit limits. But the interpretation is that the permit limit can be set without regard to the target in the receiving stream. You can check this, as noted above, by calculating the river quality downstream of various values of the discharge quality.

Occasionally you may find that you have set up a calculation which combines an upstream river quality which is worse than the target with enormous dilution of the discharge flow by the river. Here the calculated permit limit can be very sensitive to data although the downstream river quality will be almost unaffected by the discharge.

In practice the permit limit can be set without regard to the river target. Again you can check this point by calculating the effect on water quality of the current discharge load. Discharges with Little Dilution

When the 5-percentile river flow is zero or some small proportion of the discharge flow, the calculated permit limit will come out close to the concentration entered for the river target. This reflects the fact that where there is no dilution from the river, the distribution of discharge quality must be the equivalent of that required for the river target.

If this is unacceptable, we will either have to live with failure to meet the river target or insist that the discharge goes elsewhere.

Some decisions of this type depend on whether a small watercourse is important enough to be treated as a river and so be classified under the Water Framework Directive.


Supporting Calculations Standard Deviation from Mean and Percentile

You can do these calculations using RQP by clicking on “CALCULATION OF MEAN AND STANDARD DEVIATION FROM A PERCENTILE” on the initial RQP screen (see page 22 ‘Doing the calculations’).

You enter in the blue boxes the percentile concentration, the percentile point (for example 90 for the 90-percentile), and your estimate of the coefficient of variation (the ratio of the standard deviation to the mean). The calculated mean and standard deviation appears in the white boxes.

As shown below in the copy of the screen, values of 4, 90 and 0.7 for the input data, give a mean of 2.17 and a standard deviation of 1.52.

Your estimate of the coefficient of variation should be based data on river quality though it is usually safe to use the following:

BOD 0.6 Phosphate 0.6 Ammonia 1 Metals 2


Calculation of Correlation Coefficients

One of the programmes on the RQP screen allows you to calculate correlation coefficients. To use it click on “CALCULATION OF CORRELATION COEFFICIENTS”.

You might want to do this if you have found a calculation in which the decision on the permit limit is sensitive to the value of one or more of the correlation coefficients.

You will be asked you for the name of a data-file. An example, called REG.DAT, is provided as part of RQP. You double-click on the file name and then click calculate”.


The data-file consists of pairs of values of the variables for which you want to calculate the correlation coefficient. REG.DAT has the 12 rows of data shown below. In the first row, for example, the values 2 and 4, could be corresponding values of discharge flow and river flow - values measured at the same time.

2 4 3 6 2 5 4 8 1 9 3 5 2 8 10 20 3 5 2 2 1 2

0.1 0.7

The programme takes logs of these data and displays the correlation coefficient between the logged variables. Discharges to Tributaries

Some discharges enter a river close to the confluence with another watercourse. Here, you may need to take account of the target quality in both each rivers.

It may be important to allow for natural purification in the river which first receives the discharge and check what is needed to meet the target in the second river. Alternatively, you can use SIMCAT for this type of calculation.

Appendix B


Appendix B Calculating Permit Limits for Priority or Hazardous Substances

METHOD FOR SETTING PERMIT LIMITS FOR PRIORITY AND HAZARDOUS SUBSTANCES AND ASSESSING COMPLIANCE

For all discharges that are liable to contain Priority or Hazardous Substances we will use one of four control regimes according to the risk from each substance to the environment. Each control regime has a different set of Standard Conditions in the Permit. Control regimes Regime A is for the most environmentally significant discharges. We will use this regime to control the discharge of any substances that could threaten compliance with any Environmental Quality Standards, Directives or international obligations, or our regulatory approach on “No Deterioration”. We will set specific numeric permit limits and ensure that these are met by requiring monitoring of the effluent. Regimes B, C and D will be applied where a discharge is assessed as liable to contain priority or hazardous substances but at such low levels that the discharge could not (by itself or combined with other discharges) threaten the achievement of any Environmental Quality Standards, Directives or international obligations, or our regulatory approach on “No Deterioration”. Regime B controls Priority Substances. We will require compliance with general standards. Regime C controls Specific Pollutants that have EQSs set by Tables 1 to 20 of the River Basin Districts Typology, Standards and Groundwater threshold values (Water Framework Directive) (England and Wales) Directions 2009. We will not set numeric limits in regime C permits. Regime D applies to other substances with non-statutory Environmental Quality Standards (Operational Standards). We will not set numeric limits in regime D permits. Numeric Permit Limits for Regime A Two numeric limits need to be met to ensure protection of the environment:

a) A lower tier limit to be met for at least 95% of the time and assessed annually using an appropriate look up table;

b) An upper tier limit set at an agreed multiple of the middle tier limit, to be met all of the time.

For regulation we set the lower tier and the upper tier as discharge Permit limits. The lower tier is set at the 95%ile value with compliance assessed by look up table. The upper tier is set at twice the 99.5%ile value. .


The main control on the substance is through the 95%ile limit. The look up table provides for sampling variation to give substantial protection from false failures if the discharge does not deteriorate. The upper tier allows enforcement in case of a significant short term pollution event and the Dangerous Substances Directive requires maximum limits for List I substances. The method of calculation is described in simple terms below. When will we apply Regime A numeric limits? Regime A is used when the discharge meets at least one of the following 5 criteria: 1. Failure of any sample to comply with the General Standards The General Standards are found in Appendix A of Annex D. If any sample of the discharge exceeds these values we will use Regime A. 2. Threat to the Environmental Quality Standard The RQP software (see Appendix A) will be used to calculate the downstream concentration of the substance. If this shows more than 95% confidence that the EQS will be exceeded, we will use Regime A. 3. Risk of Significant Deterioration in Receiving Water Quality If the calculated downstream concentration shows an increase over the upstream concentration which is more than 10% of the EQS, we will use Regime A. 4. Significant load For a substance in the following table, if the calculated load exceeds the value in the table, we will use Regime A.

Table 4.1 Annual significant loads for priority and hazardous substances 1,2-dichloroethane 5kg/y aldrin 30g/y arsenic 20kg/y atrazine 100g/y cadmium 5kg/y carbon tetrachloride 1.5kg/y chloroform 40 kg/y dieldrin 50g/y endrin 30g/y hexachlorobenzene 40g/y


hexachlorobutadiene 30g/y

mercury 0.5kg/y para-para-DDT 90 g/y pentachlorophenol 250g/y perchloro (tetrachloro) ethylene

0.5kg/y

simazine 5g/y trichlorobenzene 200g/y trichloroethylene 5kg/y

5. Unacceptable Risk of Deterioration in Effluent Quality If a treatment plant is required specifically to remove a priority or hazardous substance we will use Regime A. If a priority or hazardous substance is added as part of a treatment process, we will use Regime A. If this is a new discharge or there is no historic data take the upstream flow and concentration data, together with discharge flow, and predicted input concentration to calculate the effect of input discharge quality downstream using MCARLO or SIMCAT (see appendix A). You may use a calculated upstream flow if the real data is not available, and if real concentration data is not available assume 10% of EQS where there are no upstream inputs of the substance, and 50% of EQS where there are known inputs. If the increase in concentration from upstream to downstream is less than 10% of the EQS, then we will use Regime B, C or D. If not, Regime A will be used in the permit. We will use MCARLO or SIMCAT to calculate the required discharge quality (as a 95%ile and 99.5%ile) for the discharge which would give an increase of 10% of the EQS from upstream to downstream. If you believe that you cannot meet this discharge quality as a permit limit, you need to contact us. If you have historic data for the discharge then follow the HiTail method below to calculate the ratio of 99.5%ile to the Base%ile. If you do not have any suitable data for the site, or it is a new discharge, then use the PCCut = 40% tables below to read off the 99.5%ile : 95%ile ratio for your determinands. Having calculated the ratio apply it to the calculated 95%ile to give a 99.5%ile. Double this value and set it as the upper tier limit.


Calculating Numeric Permit Limits when you have historic data, using the HiTail Method Step Task

1 If you have a historical data set for the determinands of interest for the

discharge, follow the steps below for each of the determinands. 2 Replace all less-than values by less-than the MRV. Then plot the data as a

time series. If there is an obvious trend, consider removing early years as necessary until the data represents a period of broadly stable quality reflecting the most recent situation.

3 To cater for datasets that will often have a proportion of ‘less-than’ values this method only uses the upper portion of the dataset that contains only real values. This avoids questionable assumptions about the treatment of ‘less-than’ values in calculations. The discarded ‘less-than’ values are allowed for in the method. The portion of the dataset containing the ‘less-than’ values that is discarded is termed the PCCut value. Tables are provided for PCCut values of 40%, 60%, 80%, 90% and 95%. First, calculate the percentage of less-than values. If this is greater than 95%, abandon the exercise as you have too little real data. Otherwise, round the percentage up to the next PCCut value for which tables are provided. That is the lowest PCCut % that has no less-than values. For instance, if the dataset has 75% of the data as ‘less-than’ values you should use a PCCut value of 80%. If there are no less-than values assume a PCCut of 40%

4 Sort the data into order from lowest to highest value and remove the selected PCCut %. This leaves the remaining (100 - PCCut)% of values in what we term the HiTail.

5 Take logs to base 10 of the HiTail values and calculate their standard deviations.

6 Consult the SD Factor table for the PCCut value that you have used (Tables 1-5) and locate the entry for your determinand. If there is not an entry in the table for your determinand check Tables 6 and 8 to decide the appropriate default value for the SD Factor.

7 The SD Factor tables have three columns for different ratios of the 99.5%ile to the Base%ile. The Base%ile is the percentile from which you will calculate the 99.5%ile. You should use the lowest possible percentile as the Base%ile. If your PCCut value is 40%, 60% or 80% then use the 80%ile as the Base%ile. If the PCCut value is 90% then use the 90%ile as the Base%ile and if the PCCut value is 95% then use the 95%ile. Read off the SD Factor value, k, for the Base%ile you are using.

8 You should then calculate the ratio of the 99.5%ile to your Base%ile. Use the SD Factor k (from Step 7) and the standard deviation of the log values, s (from Step 5). The formula is: Ratio of (99.5%ile / Base%ile) = 10(k×s)

This gives a percentile ratio that is tuned to the specific substance under consideration.


9 Take the sorted data identified in step 4 without the HiTail removed. Take the Base%ile value from the sorted data. The Base%ile is the ‘rth’ sample value from lowest to highest, where: r = (Base%ile/100)×(N+1) Where N is the number of values in the data set. When r has a fractional component (e.g. 45.7) then the Base%ile should be determined by interpolation from the adjacent values (e.g. If the result is 45.7, then calculate 0.7 times the difference between the 45th and 46th values and add the result to the 45th value). Note that this method will always produce a real data value, because your PCCut value is always less than your Base%ile (from Step 3).

10 If the Base%ile is not the 95%ile, then you should also calculate the 95%ile using the method in Step 9. This will be set as the 95%ile limit for a ‘standstill’ Permit limit.

11 Finally, calculate the 99.5%ile as: 99.5%ile / Base%ile ratio (from step 8) × Base%ile (from step 9) then double the 99.5%ile to give the upper tier limit that you will use in the Permit.

GLOSSARY PCCut – The percentage of values needing to be discarded from the ranked sample data for any determinand to ensure that the remaining data contains no less-than values. HiTail – The HiTail is the part of the dataset that is left after the PCCut has been removed. It will only contain real values. This is also the name given to this method developed by WRc for deriving extreme percentile values from three elements: a Base%ile; the standard deviation of the logged data above the PCCut and the relevant SD Factor. SD Factor – A number that when multiplied by the standard deviation of the logged data in the HiTail and antilogged gives the ratio between two specific percentiles. Base%ile – A percentile (e.g. 95%ile) that can be readily estimated from a historical dataset and from which the 99.5%ile can be estimated.


Table 1 - SD Factor values for PCCut = 40%

SD Factor for: 99.5/80 99.5/90 99.5/95 Determinand

3.2 2.6 2.0 Copper - as Cu 2.4 1.7 1.3 Copper dissolved - as Cu 3.9 3.2 2.3 Iron - as Fe 3.1 2.4 1.6 Iron dissolved - as Fe 3.4 2.7 2.1 Lead - as Pb 2.6 1.7 1.3 Lead dissolved - as Pb 4.2 3.4 2.8 Mercury - as Hg 3.4 2.8 2.1 Nickel - as Ni 4.1 3.7 2.0 PCP and compounds 3.0 2.4 1.7 tetrachloroethane (per/tet) 4.0 3.4 2.8 Trichloromethane (chloroform) 3.5 2.7 2.1 Zinc - as Zn 3.0 2.4 1.8 Zinc dissolved - as Zn

Table 2 - SD Factor values for PCCut = 60%

SD Factor for: 99.5/80 99.5/90 99.5/95 Determinand

3.8 3.0 2.5 Cadmium - as Cd 3.5 2.7 2.0 Copper - as Cu 2.7 1.8 1.4 Copper dissolved - as Cu 3.7 3.0 2.1 Iron - as Fe 3.0 2.3 1.4 Iron dissolved - as Fe 3.5 2.7 1.9 Lead - as Pb 3.0 2.0 1.5 Lead dissolved - as Pb 3.8 2.9 2.1 Mercury - as Hg 3.6 2.9 2.2 Nickel - as Ni 3.9 3.1 2.4 Nickel dissolved - as Ni 3.4 3.0 1.7 PCP and compounds 3.1 2.3 1.5 Tetrachloroethene (per/tet 3.6 3.0 2.0 Trichloroethene (trichloro 4.0 3.4 2.6 Trichloromethane (chloroform) 3.7 2.9 2.2 Zinc - as Zn 3.3 2.6 1.8 Zinc dissolved - as Zn


Table 3 - SD Factor values for PCCut = 80% SD Factor for:

99.5/80 99.5/90 99.5/95 Determinand 3.6 3.0 2.4 Cadmium - as Cd 3.7 3.0 2.1 Copper - as Cu 3.7 2.5 1.5 Copper dissolved - as Cu 4.0 3.3 2.5 Dichlorvos 4.0 3.3 2.5 HCH gamma 3.5 2.8 1.9 Iron - as Fe 3.0 2.2 1.3 Iron dissolved - as Fe 3.5 2.7 1.8 Lead - as Pb 3.8 2.6 1.7 Lead dissolved - as Pb 3.7 3.0 2.1 Mercury - as Hg 3.7 3.0 2.3 Nickel - as Ni 4.0 3.3 2.4 Nickel dissolved - as Ni 2.8 2.5 1.3 PCP and compounds 3.6 2.8 1.9 Tetrachloroethene (per/tet) 3.9 3.2 2.0 Trichloroethene (trichloro) 3.6 3.2 2.4 Trichloromethane (chloroform) 3.7 3.0 2.2 Zinc - as Zn 3.5 2.7 1.9 Zinc dissolved - as Zn


Table 4 - SD Factor values for PCCut = 90% SD Factor for:

99.5/80 99.5/90 99.5/95 Determinand 2.9 2.2 Atrazine 3.7 2.2 Cadmium - as Cd 3.2 2.4 Copper - as Cu 3.0 1.8 Copper dissolved - as Cu 3.1 2.3 Dichlorvos 3.1 2.3 HCH gamma 3.0 2.3 Iron - as Fe 3.0 2.2 Iron dissolved - as Fe 3.1 2.2 Lead - as Pb 3.2 2.0 Lead dissolved - as Pb 3.0 2.3 Mercury - as Hg 3.1 2.3 Mercury dissolved - as Hg 3.0 2.4 Nickel - as Ni 3.0 2.4 Nickel dissolved - as Ni 2.7 1.6 PCP and compounds 3.1 2.2 Simazine 3.1 2.2 Tetrachloroethene (per/tet) 3.2 2.2 Trichloroethene (trichloro) 3.2 2.4 Trichloromethane {chloroform) 3.1 2.3 Zinc - as Zn 3.2 2.3 Zinc dissolved - as Zn


Table 5 - SD Factor values for PCCut = 95% SD Factor

99.5/80 99.5/90 for 99.5/95 Determinand 2.8 1.1.1-trichloroethane 3.0 1.2-dichloroethane (Ethylene

Dichloride) 2.8 Atrazine 3.3 Cadmium - as Cd 2.4 Copper - as Cu 2.3 Copper dissolved - as Cu 3.0 Dichlorvos 2.8 Dieldrin 3.0 HCH gamma 2.4 Iron - as Fe 5.5 Iron dissolved - as Fe 2.9 Lead - as Pb 2.5 Lead dissolved - as Pb 2.7 Mercury - as Hg 2.5 Mercury dissolved - as Hg 2.5 Nickel - as Ni 2.6 Nickel dissolved - as Ni 2.7 PCP and compounds 3.1 Simazine 2.5 Tetrachloroethene (per/tet 2.5 Trichloroethene (trichloro 2.8 Trichloromethane (chloroform) 2.7 Zinc - as Zn 3.8 Zinc dissolved - as Zn


SD Factors for Substances not in the above Tables The R&D Report analysed a representative range of substances. For other substances it recommends that substances are classified on seasonality and whether they tend to be in the dissolved or particulate phase in treated discharges. You should therefore decide from Table 6 into which class your substance of interest falls. Some of these classes are further divided into different chemical groups and Table 7 gives examples of some of the substances that fall into each group. Having decided into which Group your substance falls then read off the SD Factor from the appropriate point in Table 8. Table 6 – Substance classification Group Substance

Class Description

1. Total metals 2. Dissolved metals 3. Solvents

A. Non-seasonal dissolved

These are substances that would be likely to be present in sewage works effluent at a consistent level because inputs and removal are broadly constant. Consistent inputs might arise as a result of many diffuse inputs (e.g. copper and zinc from domestic sources), a single consistent source (e.g. fluoride dosing of drinking water), lack of seasonality of use (various industrial chemicals). The fact that these substances are also predominantly associated with the dissolved phase in effluents would tend to lead to consistency in sewage treatment and a low CoV. (Examples copper, zinc, nickel, chromium, trichloroethane, trihalomethanes).

4. Total metals 5. Dissolved metals

B. Non-seasonal particulate

This class is as A, above, with respect to occurrence, but would include substances that predominantly are associated with the particulate phase in effluents. The concentrations and CoV values for these substances would be more clearly influenced by works' performance – or, at least, the effectiveness of solids removal. Hence CoV values would tend to be higher at a treatment works where solids concentrations varied, and between works. (Examples: aluminium, iron, lead, BDEs, manganese APEs, APEOs).

6. Biocides, etc

C. Seasonal dissolved

These are determinands that are discharged or used on a seasonal or ad hoc basis and are predominantly present as dissolved species in the effluent. This would lead to a high annual CoV that might be similar across areas or different treatment works. (Examples: atrazine, simazine, other relatively soluble herbicides, DEHP (in car washes), soluble substances that are present owing to an intermittent local discharge).


7. Biocides, etc

D. Seasonal particulate

This class is as C, but includes substances associated with particulates. These substances would be subject to a high annual CoV and be less comparable with respect to CoV across treatment works (examples PAHS, pesticides with high Kow, high Kow substances present owing to an intermittent local discharge).

Table 7 – Examples of substances in each Group Group Class Example substances 1. Total metals A Antimony \ Arsenic \ Molybdenum \ Nickel \ Selenium 2. Dissolved metals

A Antimony \ Arsenic \ Molybdenum \ Nickel \ Selenium

3. Solvents A 1,1,2-Trichloroethane \ 1,2-Dimethylbenzene (O-Xylene) \ Benzene \ Ethenylbenzene {Vinylbenzene} {Styrene} \ Tetrachloromethane (Carbon Tetrachloride) \ Toluene (Methylbenzene) \ Xylene (m and p)(1,3+1,4-Dimethylbenzene)

4. Total metals B Chromium (VI) \ Methylmercury \ Platinum group metals \ Silver

5. Dissolved metals

B Cadmium \ Chromium (VI) \ Methylmercury \ Platinum group metals \ Silver

6. Biocides, etc C Fenitrothion \ Fenthion \ Malathion \ Parathion (Parathion Ethyl) \ Parathion-Methyl \ DEHP

7. Biocides, etc D Aldrin \ DDT (op) \ DDT (PP) \ PAHs


Table 8 – SD factor values for other substances SD Factor values for PCCut = 40%

Group (from Table 6) 99.5/80 99.5/90 99.5/95 1 3.4 2.7 2.1 2 3.0 2.4 1.8

3 3.7 3.2 2.2

4 3.7 3.0 2.3 5 2.7 1.9 1.4

6 3.7 3.0 2.5

7 2.6 2.3 1.3

SD Factor values for PCCut = 60%

Group (from Table 6) 99.5/80 99.5/90 99.5/95 1 3.6 2.9 2.2

2 3.6 2.9 2.1 3 3.6 2.9 2.0

4 3.7 2.9 2.1

5 2.9 2.0 1.4 6 3.9 3.2 2.4

7 3.4 3.0 1.7


Group (from Table 6) 99.5/80 99.5/90 99.5/95 1 3.7 3.0 2.2 2 3.7 3.0 2.1

3 3.7 3.1 2.1

4 3.6 2.9 2.0 5 3.5 2.4 1.5

6 4.0 3.3 2.5

7 2.8 2.5 1.3



Group (from Table 6) 99.5/80 99.5/90 99.5/95

1 3.0 2.4

2 3.1 2.4

3 3.2 2.3 4 3.2 2.3

5 3.1 2.1

6 2.1 2.3

7 2.7 1.6


Group (from Table 6) 99.5/80 99.5/90 99.5/95 1 2.6

2 3.2 3 2.7

4 2.7

5 2.5 6 3.0

7 2.8


WORKED EXAMPLES – CALCULATING PERMIT LIMITS FOR REGIME A EXAMPLE 1 Setting the limits Data from three years sampling for zinc is shown in Table 1. The means for these years are 24.6 μg/l, 25.8 μg/l and 23.6 μg/l, suggesting that the values are consistent over these three years. TABLE 1 – Historic values for zinc

Date Value Unit Date Value Unit 10-Jan-00 30.0 μg/l Zn 06-Aug-01 24.6 μg/l Zn 07-Feb-00 31.0 μg/l Zn 03-Sep-01 31.2 μg/l Zn 07-Mar-00 23.0 μg/l Zn 24-Sep-01 23.9 μg/l Zn 07-Apr-00 20.0 μg/l Zn 23-Oct-01 22.7 μg/l Zn

15-May-00 28.0 μg/l Zn 13-Dec-01 30.0 μg/l Zn 13-Jun-00 30.0 μg/l Zn 11-Jan-02 27.1 μg/l Zn 29-Jun-00 32.0 μg/l Zn 06-Feb-02 16.1 μg/l Zn 08-Aug-00 20.0 μg/l Zn 06-Mar-02 26.9 μg/l Zn 06-Sep-00 30.0 μg/l Zn 08-Apr-02 24.6 μg/l Zn 12-Oct-00 16.0 μg/l Zn 08-May-02 12.8 μg/l Zn 07-Nov-00 12.0 μg/l Zn 29-May-02 10.8 μg/l Zn 01-Dec-00 23.0 μg/l Zn 08-Jul-02 13.9 μg/l Zn 29-Jan-01 23.0 μg/l Zn 31-Jul-02 20.7 μg/l Zn 07-Feb-01 20.0 μg/l Zn 29-Aug-02 32.7 μg/l Zn 23-May-01 27.8 μg/l Zn 26-Sep-02 31.7 μg/l Zn 19-Jun-01 31.8 μg/l Zn 21-Oct-02 33.4 μg/l Zn 18-Jul-01 23.1 μg/l Zn 20-Nov-02 32.0 μg/l Zn

Sorting the data values from lowest to highest gives the following dataset: 10.8, 12.0, 12.8, 13.9, 16.0, 16.1, 20.0, 20.0, 20.0, 20.7, 22.7, 23.0, 23.0, 23.0, 23.1, 23.9, 24.6, 24.6, 26.9, 27.1, 27.8, 28.0, 30.0, 30.0, 30.0, 30.0, 31.0, 31.2, 31.7, 31.8, 32.0, 32.0, 32.7, 33.4 There are 34 data values and the 95%ile value is calculated from the formula: r = P(n + 1), or r = 0.95(34 + 1) = 33.25 That is the 95%ile value lies between the 33rd and the 34th data point. The difference between these two points is 0.7 μg/l (i.e. 33.4 – 32.7). Multiply that by the fractional part of r (i.e. 0.25) and add the result to the 33rd value gives 32.9 (i.e. (0.25 x 0.7) + 32.7). For regulatory purposes, we would normally round that up to a 95%ile Permit limit of 33 μg/l.


The 99.5%ile value is calculated by the method given in Appendix 1 above. We set the PCCut as 40% because there are no less than values. We therefore use the 80%ile as the Base%ile. The three elements for calculation are: • the Base%ile of the dataset = 31.2 μg/l; • the SD Factor (k) =3.5 for the 80%ile/99.5%ile ratio for total zinc; • The standard deviation, s, for the logged data is 0.135. This gives a P99.5/P80 ratio of 2.388 times the 80%ile for this dataset, which is a 99.5%ile value of 74.5 μg/l. Multiplying that by 2 gives an upper tier Permit limit of 149 μg/l. Compliance Applying these limits to the results from the following three years, we get the results in Table 2. TABLE 2 – Monitoring values for zinc

Date Value Unit Date Value Unit 24-Jan-03 28.0 μg/l Zn 24-Aug-04 20.2 μg/l Zn 10-Feb-03 27.5 μg/l Zn 21-Sep-04 26.2 μg/l Zn 12-Mar-03 24.1 μg/l Zn 14-Oct-04 26.5 μg/l Zn 09-Apr-03 40.2 μg/l Zn 10-Nov-04 22.7 μg/l Zn

07-May-03 25.8 μg/l Zn 03-Dec-04 26.5 μg/l Zn

04-Jun-03 17.7 μg/l Zn 28-Jan-05 24.9 μg/l Zn 15-Sep-03 29.6 μg/l Zn 21-Mar-05 22.7 μg/l Zn 10-Oct-03 25.2 μg/l Zn 12-Apr-05 23.7 μg/l Zn 28-Oct-03 25.6 μg/l Zn 17-May-05 15.1 μg/l Zn 06-Nov-03 26.8 μg/l Zn 08-Jun-05 13.0 μg/l Zn 12-Dec-03 31.3 μg/l Zn 01-Jul-05 8.8 μg/l Zn

12-Jan-04 28.9 μg/l Zn 31-Aug-05 18.3 μg/l Zn 20-Feb-04 26.3 μg/l Zn 20-Sep-05 27.5 μg/l Zn 24-Mar-04 32.1 μg/l Zn 12-Oct-05 21.7 μg/l Zn 12-May-04 20.9 μg/l Zn 21-Oct-05 17.5 μg/l Zn 02-Jun-04 30.1 μg/l Zn 17-Nov-05 25.0 μg/l Zn 05-Jul-04 25.5 μg/l Zn 05-Dec-05 23.3 μg/l Zn 28-Jul-04 29.6 μg/l Zn

In the three years, there are 1, 0 and 0 failures. Three sample failures are required in any year before the limit is failed. The discharge therefore complies in all years. The data for both periods is shown in Figure 1.


FIGURE 1 – Plot of example 1 data over both data periods

EXAMPLE 2 Setting the limits Table 3 shows three years of monitoring results for zinc in a different discharge. There is some variability over the period with successive annual means of 27.3 μg/l, 19.7 μg/l and 43.5 μg/l, with a notably high value of 108 μg/l in the last year. TABLE 3 – Historic values of zinc

Date Value Unit Date Value Unit 01-Feb-00 29.0 μg/l Zn 04-Jul-01 19.2 μg/l Zn 18-Feb-00 34.0 μg/l Zn 01-Aug-01 17.7 μg/l Zn 27-Mar-00 28.0 μg/l Zn 31-Aug-01 22.8 μg/l Zn 05-May-00 32.0 μg/l Zn 03-Oct-01 14.5 μg/l Zn 06-Jun-00 20.0 μg/l Zn 09-Nov-01 23.4 μg/l Zn 26-Jun-00 21.0 μg/l Zn 30-Nov-01 22.2 μg/l Zn 13-Jul-00 17.0 μg/l Zn 09-Jan-02 22.1 μg/l Zn

16-Aug-00 23.0 μg/l Zn 04-Feb-02 34.4 μg/l Zn 05-Oct-00 24.0 μg/l Zn 07-Mar-02 26.3 μg/l Zn 19-Oct-00 19.0 μg/l Zn 16-Apr-02 22.6 μg/l Zn 15-Nov-00 18.0 μg/l Zn 09-May-02 40.4 μg/l Zn 06-Dec-00 28.0 μg/l Zn 11-Jun-02 32.2 μg/l Zn 15-Jan-01 21.0 μg/l Zn 05-Jul-02 35.9 μg/l Zn 08-Feb-01 24.0 μg/l Zn 15-Aug-02 24.6 μg/l Zn 22-Mar-01 20.4 μg/l Zn 10-Sep-02 45.7 μg/l Zn 19-Apr-01 23.1 μg/l Zn 03-Oct-02 27.6 μg/l Zn

16-May-01 22.3 μg/l Zn 30-Oct-02 108.0 μg/l Zn 15-Jun-01 24.7 μg/l Zn 02-Dec-02 25.5 μg/l Zn

0.05.0

10.015.020.025.030.035.040.045.0

10/0

1/20

00

10/0

7/20

00

10/0

1/20

01

10/0

7/20

01

10/0

1/20

02

10/0

7/20

02

10/0

1/20

03

10/0

7/20

03

10/0

1/20

04

10/0

7/20

04

10/0

1/20

05

10/0

7/20

05

zinc

as

ug/l


This gives the ranked dataset: 14.5, 17.0, 17.7, 18.0, 19.0, 19.2, 20.0, 20.4, 21.0, 21.0, 22.1, 22.2, 22.3, 22.6, 22.8, 23.0, 23.1, 23.4, 24.0, 24.0, 24.6, 24.7, 25.5, 26.3, 27.6, 28.0, 28.0, 29.0, 32.0, 32.2, 34.0, 34.4, 35.9, 40.4, 45.7, 108.0 There are 36 data values and the 95%ile value is calculated from the formula: r = P(n + 1), or r = 0.95(36 + 1) = 35.15 That is the 95%ile value lies between the 35th and the 36th data point. The difference between these two points is 62.3 μg/l (i.e. 108.0 – 45.7). Multiply that by the fractional part of r (i.e. 0.15) and add the result to the 35th value gives a 95%ile of 55.0 (i.e. (0.15 x 62.3) + 45.7). That is a 95%ile Permit limit of 55 μg/l. The 99.5%ile value is calculated by the method given in Appendix 1 above. Since the PCCut is 40% we use the 80%ile as the Base%ile. The three elements for calculation are:

• the Base%ile of the dataset = 32.1 μg/l; • the SD Factor (k) = 3.5 for the 80%ile/99.5%ile ratio for total zinc; • The standard deviation, s, for the logged data is 0.152 (and 0.109 with the 108 μg/l

value removed, see below) This gives a P99.5/P80 ratio of 2.664 times the 80%ile for this dataset, which is a 99.5%ile value of 85.5 μg/l. Multiplying that by 2 gives an upper tier Permit limit of 171 μg/l. However, the 95%ile limit is influenced by the high value of 108 μg/l. In order to illustrate the effect of this value the limit was assessed again after removing the value from the dataset. This gave a 95%ile of 41.2 μg/l, which we round up to a 95%ile Permit limit of 42 μg/l. For the dataset without the 108 μg/l result the P99.5/P80 ratio is 2.019 giving a value for the 99.5%ile of 64.8 μg/l and an upper tier Permit limit of 130 μg/l. Compliance Applying these limits to the results from the following three years, we get the results in Table 4. Failures against the 55 μg/l limit are highlighted in yellow. Failures against the 42 μg/l limit are marked by a square. TABLE 4 – Monitoring values for zinc

Date Value Unit Date Value Unit 21-Jan-03 41.2 μg/l Zn 13-Aug-04 34.0 μg/l Zn 19-Feb-03 38.1 μg/l Zn 24-Aug-04 11.6 μg/l Zn 17-Mar-03 35.1 μg/l Zn 10-Oct-04 34.2 μg/l Zn 01-May-03 40.5 μg/l Zn 18-Nov-04 58.4 μg/l Zn 21-May-03 27.8 μg/l Zn 02-Dec-04 16.9 μg/l Zn

02-Jul-03 35.0 μg/l Zn 08-Dec-04 44.5 μg/l Zn


22-Jul-03 28.1 μg/l Zn 31-Jan-05 49.5 μg/l Zn 19-Aug-03 33.7 μg/l Zn 07-Mar-05 27.1 μg/l Zn 18-Sep-03 49.6 μg/l Zn 23-Mar-05 39.8 μg/l Zn 15-Oct-03 55.2 μg/l Zn 29-Apr-05 17.4 μg/l Zn 04-Nov-03 31.6 μg/l Zn 23-May-05 47.9 μg/l Zn 04-Dec-03 50.9 μg/l Zn 06-Jul-05 14.2 μg/l Zn

19-Jan-04 55.6 μg/l Zn 20-Jul-05 34.7 μg/l Zn 17-Feb-04 34.4 μg/l Zn 02-Sep-05 28.3 μg/l Zn 04-Mar-04 46.0 μg/l Zn 31-Oct-05 14.2 μg/l Zn 05-Apr-04 33.3 μg/l Zn 23-Nov-05 17.6 μg/l Zn

21-May-04 37.8 μg/l Zn 06-Dec-05 24.2 μg/l Zn 17-Jun-04 26.7 μg/l Zn 13-Dec-05 17.3 μg/l Zn 30-Jul-04 15.8 μg/l Zn

In the three successive years, there are 1, 2 and 0 exceedences of the 55 μg/l limit. Three sample failures are required in any year before the limit is reported as failed so against that limit the site passed in each year. The picture would have been different if the 108 μg/l had not been recorded or had been excluded from compliance assessment. Against the 42 μg/l limit the failures in each year are 3, 4 and 2. That is a Permit failure in each of the first two years. The reason for this is obvious if we examine the plot of the results over the whole six years shown in Figure 2. FIGURE 2 – Plot of Example 2 over both data periods.

It is evident that the effluent deteriorated after 2002 and the concentrations in the second part of the six years when compliance was assessed were higher than during the first part used to calculate the standstill limit. This visual impression was confirmed by comparing the two datasets using the Mann-Whitney test. The Null hypothesis was rejected at the 1% significance level. That is the difference is significant and the deterioration is real. It

0.0

20.0

40.0

60.0

80.0

100.0

120.0

01/0

2/20

00

01/0

6/20

00

01/1

0/20

00

01/0

2/20

01

01/0

6/20

01

01/1

0/20

01

01/0

2/20

02

01/0

6/20

02

01/1

0/20

02

01/0

2/20

03

01/0

6/20

03

01/1

0/20

03

01/0

2/20

04

01/0

6/20

04

01/1

0/20

04

01/0

2/20

05

01/0

6/20

05

01/1

0/20

05

Zinc

ug/

l


is not known what the cause of the deterioration was, but it illustrates the power of the method in detecting real changes. EXAMPLE 3 Date LT Hg as

ug/l Date LT Hg as

ug/l 01-Feb-00 < 0.01 01-Aug-01 0.07 26-Jun-00 < 0.01 15-Aug-02 0.07 13-Jul-00 < 0.01 04-Dec-03 0.07 19-Oct-00 < 0.01 27-Mar-00 0.08 15-Jan-01 < 0.01 06-Jun-00 0.08 08-Feb-01 < 0.01 16-May-01 0.08 22-Mar-01 < 0.01 04-Mar-04 0.08 15-Jun-01 < 0.01 13-Aug-04 0.08 31-Aug-01 < 0.01 06-Dec-05 0.08 03-Oct-01 < 0.01 17-Jun-04 0.10 30-Nov-01 < 0.01 23-Mar-05 0.10 21-Jan-03 < 0.01 18-Sep-03 0.11 30-Jul-04 < 0.01 06-Dec-00 0.12

02-Dec-04 < 0.01 09-May-02 0.14 07-Mar-05 < 0.01 02-Dec-02 0.15 29-Apr-05 < 0.01 22-Jul-03 0.15 06-Jul-05 < 0.01 03-Oct-02 0.16 31-Oct-05 < 0.01 19-Feb-03 0.16 23-Nov-05 < 0.01 31-Jan-05 0.22 13-Dec-05 < 0.01 16-Apr-02 0.25 19-Apr-01 0.01 10-Sep-02 0.25 15-Nov-00 0.02 20-Jul-05 0.25 04-Jul-01 0.02 19-Aug-03 0.27

09-Jan-02 0.02 15-Oct-03 0.33 07-Mar-02 0.02 21-May-04 0.37 24-Aug-04 0.02 30-Oct-02 0.38 17-Feb-04 0.03 19-Jan-04 0.43 18-Nov-04 0.03 09-Nov-01 0.71 05-May-00 0.04 05-Jul-02 0.89 16-Aug-00 0.04 02-Jul-03 0.89 04-Feb-02 0.04 11-Jun-02 0.90 05-Apr-04 0.04 05-Oct-00 1.00 18-Feb-00 0.05 21-May-03 1.64


10-Oct-04 0.05 01-May-03 2.50 08-Dec-04 0.05 04-Nov-03 2.67 02-Sep-05 0.05 17-Mar-03 9.40 23-May-05 0.06 Step 3: 27.4% of values are ‘less thans’ so lowest PCCut is 40%. Step 4: Lowest 40% (blue highlighted) values are removed. Step 5: Logging the remaining values and calculating the standard deviation (s)

gives a value of 0.5630 Step 6 + 7: Since PCCut ≤80% we use the 80%ile as a base to calculate the 99.5%ile.

The ratio of 99.5%ile to 80%ile for mercury for the 40% PCCut value is 4.2. Step 8: Ratio from formula is 231 Step 9: The 80%ile value is the 59.2th value, which is 0.25μg/l. The 95%ile value is

the 70.3th value, which is 1.90 μg/l. Step 10: The calculated 99.5%ile is therefore 231 x 0.25 = 57.75 μg/l.

APPENDIX C


Appendix C Evidence to show storm overflow discharges are unsatisfactory Criteria You need to provide evidence that a discharge qualifies for inclusion against

one or more of these criteria:

i. causes significant visual or aesthetic impact due to solids, fungus ii. causes or makes a significant contribution to a deterioration in river

chemical or biological class; iii. causes or makes a significant contribution to a failure to comply with

Bathing Water Quality Standards for identified bathing waters; iv. operates in dry weather conditions; v. operates in breach of Permit conditions provided that they are still

appropriate; vi. causes a breach of water quality standards (EQS) and other EC

Directives; and/or vii. causes unacceptable pollution of groundwater.

Type of information to be considered Environment Agency records

Our records include:

Public complaints

pollution incidents (use NIRS and NIRS2)

any history of dry weather flow operation

any impact on water quality class or objectives (using for example the reasons for failure or bathing waters stewardship databases).

Modelling – by us or a water company

Riverine or marine water quality impact modelling Assess impact on water quality class or objectives. This will be of assistance to check compliance with the appropriate water quality standards. For example those associated with freshwater fish, bathing or shellfish waters directives.

Sewerage modelling Spill frequency, volume and duration information will assist with impact assessment. Spill frequency modelling can be used to identify those overflows that significantly contribute to bathing waters and shellfish waters standards failures. Survey data collected to build models may help identify problems with sewerage infrastructure, for example low weir setting giving rise to operation during dry weather


Water company records

Confirm what is already installed at the discharge structure. For example, screening if available, pass forward flow, facilities in case of emergency.

Event duration monitoring records if available

Please note it is not expected that evidence would be provided under all the categories above, but the evidence must be robust and demonstrate the discharge is unsatisfactory against the relevant qualifying criteria.

Example Where a discharge qualifies as unsatisfactory due to significant visual or aesthetic impact then in you would expect to have evidence in the form of:

photographic evidence of sewage debris or sewage fungus;

a history of public complaint;

from aesthetic surveys.

You would also expect field staff to have visited the site.

Using judgement Water quality planning, regulatory specialist and field staff should use their judgement to determine whether there is sufficient evidence to justify the inclusion of a discharge for improvement works. Identified discharges may be subject to review to determine whether there is sufficient evidence to identify them as unsatisfactory.

4 Annexes


H1 Annex E – Surface Water Discharges (complex)

Documents

Transcript of H1 Annex E – Surface Water Discharges (complex)