Skip to main content

Technical guidance

Further technical guidance

This page brings together a number of ancillary guidance documents and previous email bulletin information relating to modelling of marine fish-farms for CAR licence applications.

Auxiliary biomass sampling transect

In areas where the tidal residual current is weak, the auxiliary transect (site-specific sampling 2) may be aligned along a bearing reciprocal to the primary transect.  As a guide, a reciprocal transect may be suitable where the 30 ITI contour is more than 25m from the cage edge on the ‘upstream’ side, with respect to the residual current direction.  Otherwise, the auxiliary transect is best placed either adjacent and parallel to the primary transect, or at some (<30o) deviation from the primary transect.

Bath treatment depth

The 72hr model takes the results of the 3hr model as the starting point of its iteration; it is important to ensure that the cage depth is specified to permit at least one treatment in 3 hours, otherwise it will attempt to model a treatment of zero cages, and will fail.  It may be necessary to manually optimise the treatment depth value to the third decimal place, in order to maximise the 24hr treatment limit, e.g. treatment depth of 2.871m rather than 2.8m, this is effectively a treatment depth of 2.9m.

Biomass sampling stations

Using the AutoDEPOMOD mapping tool

There are two methods by which suitable sampling stations can be created using the mapping tool:

  • set the ‘regular transect sampling’ interval to ‘10m’, and select the two stations either side of the 30ITI contour; or,
  • having created a transect, use the measuring tool (SHIFT + 'set-sampling-station' tool Set Sampling Tool ) to identify the distance from the transect start point to the 30ITI contour, position the cursor on transect at 30ITI±10m, and release the mouse-key.  This technique is improved by zooming in on the area of interest (the transect), and by ensuring the mapping tool is expanded to the maximum extent afforded by the PC desktop.

Determining transect direction

Cage configuration, stocking plans, bathymetry and currents combine to make setting transects that will adequately provide a robust measure of the impact of farming operations something of a dark art, but the following is a guide:

  1. look at the data summary to determine predominant current axes at cage-bottom and near-bed - the sub-surface flow has little bearing on deposition except at very shallow sites;
  2. based on the relative strength of these (i.e. proportion exceeding resuspension threshold), determine either the mean of the two, or use near-bed axis;
  3. examine the relative strength of the residual to the current amplitude (along major axis); if the residual is of same order of magnitude or greater, pick some bearing between the two, otherwise use the major axis and ignore residual;
  4. see if this matches up with the modelled deposition footprint - bear in mind that the  footprint tends to elongate where it lies across steep slopes (because it travels further before it settles)
  5. tweak the bearing identified in 1 to 3, above, in the mapping tool to try and follow the depth contours if possible, i.e. aim for as a level a transect as possible.
  6. avoid placing transects that follow nebulous tongues of deposition or cross isolated peaks, or that don’t have a continuous decrease in concentration, i.e. no reversals of gradient.

Sampling transects at highly resuspensive sites

At sites where AutoDEPOMOD predicts so little deposition beneath the cages that the 30ITI boundary is less than 25m from the cages, the default sampling protocol should be employed, with sampling stations at cage-edge, 25m and 50m; the direction of the sampling transect may be determined by running AutoDEPOMOD with resuspension turned off, but the predicted ITI values at the sampling stations should be derived with resuspension on.  To achieve this , a transect can be determined whilst viewing one set of model results and then exported to a text file; subsequently, whilst viewing the output from a different model run, this transect can be imported, sampling stations set, and impact predictions derived for reporting.

Only the cage edge location (i.e. start of the transect), direction and depth should be specified in the ‘biomass’ sheet of the Marine Summary workbook; leave the AZE ±10m section blank.

Current meter data collected prior to the requirement for pressure data

Both treatment chemical and biomass licenses are now primarily derived from the current data; consequently, there is an enhanced environmental risk associated with this data.  Therefore, SEPA must insist on corroborative evidence that the current data truly represents the flow-field at the cage site.

The depth/pressure data component of the current meter record serves three functions:

  • to corroborate the current meter depth and position;
  • to corroborate the time of the deployment and;
  • to aid the repair of gaps in the data.

The first two are therefore for quality control (QC), whilst the third reduces the need to redeploy the meters.  SEPA will accept up to 6 hour of repair in each current meter dataset (see [LINK to] Attachment VIII to the ‘Fish Farm manual’).  A continuous pressure data record makes the repair easier since it is often clear from the pressure data where in the tidal cycle the repair is necessary and data from an equivalent period of the tidal cycle can then be incorporated into the dataset with a better accuracy.  It should be noted that data from sites with weak tidal currents, relative to other influences, are difficult to repair satisfactorily.

Current meter data that are not accompanied by a pressure record may be accepted if alternative satisfactory means of depth, position and time quality control are available.

Examples of alternative QC evidence include:

  • Meter serial numbers in header data.
  • Deployment log sheets with time, date, position, meter serial numbers, &c.
  • Predicted tidal heights, i.e. times of spring/neap tides and clear differentiation between spring and neap currents in the data.

SEPA’s assessment of older data will be made on a case-by-case basis, evaluating the environmental risk resulting from their use.  If in doubt, please contact FFModelling to discuss the issue further.

Far-field area calculation error

It has come to our attention that there is an error in the way in which AutoDEPOMOD calculates the far-field area.  The code in the FFMT sheet determines the area of each group based on the cage dimensions plus 100m in all planar directions; it then sums these to derive a value for the whole site.  This introduces an error where groups are less than 200m apart, since their respective far-field areas will overlap.

In many situations the error is minimal or does not affect the outcome, i.e. when the site passes without iterating to an EQS compliant treatment.  Where this is not the case, the following workarounds may be employed:

1) Estimate the area of overlap (this is easier for square cages) and subtract this value from the total area calculated in the FFMT workbook; input the new value into the appropriate cells in the Calicide and Slice EQS Test sheets of the workbook.

2) Where the site consists of two or more groups that form part of a continuous mooring grid but with gaps, for instance to accommodate a feed-barge or where an odd-numbered group has been specified using two groups, specify the whole mooring grid to derive an estimate of the area and then re-specify the individual groups; again, replace the appropriate values in the EQS Test sheets.  It may be necessary to make a slight 'guestimate' adjustment to account for the missing cages as the resulting value will still be in error, but hopefully not as much as if using the value generated for multiple groups.

3) Calculate the value manually from first principles; again, replace the appropriate values in the EQS Test sheets.

Whichever technique is employed, its use should be reported.

We are keen to hear from anybody who can propose an automated solution to this issue - please contact FFModelling.

Hydrographic data analysis tool notes


The current version of SEPA’s wind and current data analysis tool has the following features:

  • accepts current meter or wind data;
  • automated detection of temporal resolution;
  • calculates grid convergence;
  • automated identification of major axis;
  • tailored single-page summaries;
  • key outputs formatted for easy transfer to BathAuto and Marine-summary workbooks.


Please note that the magnetic variation and grid convergence corrections follow the clockwise convention for numeric bearings, whereby corrections to the east are positive and those to the west are negative, e.g. '4.5 degrees west' is entered as '-4.5' or '355.5' - in this example, failure to specify the right sign results in a 9 degree error in the resulting 'corrected' value.

Export 'summary'

The 'summary' sheet has a 'print-area' already set (this can be specified via the File menu) - if this is printed to PDF, the resulting files can be inserted into a report or circulated as an application-independent summary (and tend to be much smaller than image files).  Free PDF-printer software is available from the web; CutePDF Writer is small and has been found to be adequate for this purpose (this requires PS2PDF converter software, e.g. GhostScript, to be installed) – there are many equivalent alternatives.

Installing SURFER

It is important that Surfer is run independently at least once before using AutoDEPMOD, as there are ActiveX components that are only registered on the initial use; it is these that AutoDEPOMOD calls.  Failure to do so will produce a ‘Run-time error 429’ error code.

Limits to in-feed chemicals released from dispersive sites

Where AutoDEPOMOD modelling of in-feed chemicals results in significant export beyond the confines of the model domain, SEPA requires that an assessment is made of the potential to cause a reduction in the ecological status of the benthos.  Beyond the far-field allowable zone of effect, surrounding a fish farm, in-feed chemical concentrations must not exceed the defined environmental quality standards.  The fate of material exported from a site needs to be assessed taking account of the following:

  • the near bed current meter record, and the residual currents
  • the bathymetry and coastline of the surrounding area, in particular, whether the material is likely to be transported to a constrained or unconstrained receiving water,
  • any major potential deposition areas that the material is likely to be transported towards
  • the potential for cumulative deposition,
  • the likely concentration of the material in the receiving water in relation to the far-field EQS criteria,
  • the area of impact of the material if it is distributed evenly at the far-field EQS concentration.

In addition to the license values achieved with the modelling process, the applicant must also recommend a set of license values which consider the assessment; the latter should be presented in the Marine Summary workbook.

Read an outline procedure (84k) for assessing the potential far-field impact and for determining total allowable quantities based on available receiving area.

Shallow sites

SEPA’s hydrographic data specifications allow for reduced current meter data at sites in charted depths of less than 15m (chart datum [CD]), where only sub-surface and near-bed records are required.  However, AutoDEPOMOD requires current data from three depths.  This may be circumvented by using the sub-surface dataset as a dummy cage-bottom record, configured at some nominal height above the bed, just below (≤1m) the sub-surface record.

Site survey and model bathymetry

SEPA requires that depth is measured at the cage group corners as part of the site survey; in addition, where charted bathymetric data is sparse, further depth measurements are required at a distance from the cage group.  This is to ensure that the bathymetry of the area susceptible to the greatest impact due to particulate wastes, beneath the cages and in the immediate environs, is adequately represented in the configuration of the model.

It is therefore crucial that these data are integrated with any charted data prior to generation of the AutoDEPOMOD grid files, and that the realism afforded by the process is credible.  Both the method used to integrate the data, and an assessment of the success of its results should be described in the modelling report.

Viable biomass

Ensure that your passing model run is compliant with any target or maximum stocking density agreed in advance with the applicant.  The iteration process may lead to a compliant biomass with a higher stocking density than may be feasible, particularly if you elected to ignore the limit in the ‘Biomass limit’ dialogue by pressing ‘Cancel’.  It may be necessary to undertake a further single run, under the worst-case tidal condition, at the maximum feasible stocking density.

AutoDepomodv2 automates the iteration to a compliant discharge scenario, but care should be taken that the final answer is the result that you require.

The method described has been developed to try and encourage farming in the most suitable locations.  For a given hydrographic profile, a deeper site will perform more favourably than a shallower one.  For a given depth, a well-flushed site will perform more favourably than a quiescent one. For a given site, a cage layout which has shallow, well spread out cages, with a low stocking density will perform more favourably than a layout with deep cages, close together with a high stocking density.  These factors should all be taken into account when modelling the site.  Liaison with the applicant will be required to ensure that the final design of the cage set-up is realistic, in order to avoid a potential breach of license.

Windows 7, Office 2010 & AutoDEPOMOD

AutoDEPOMOD and the mapping tool have been tested under the Windows7 64 bit environment with Office 2010.  They have been shown to work after careful installation of the transect tool; the following describes this process:

The DLL that the transect tool uses does not register properly upon installation under Windows 7 (32 or 64 bit) - so it doesn't work – fixing this involves two factors:

  1. You need to use the version of regsvr32.exe stored in C:\Windows\SysWoW64 not the one on C:\WINDOWS\System32 (this affects only the 64 bit system)
  2. You need to register the DLL as an Administrator (this affects both 32 and 64 bit versions of Windows7)

Here's how to do it:

  1. In the new command window navigate to C:\windows\SysWoW64, i.e. type cd c:\windows\syswow64
    • type and execute: regsvr32.exe "C:\SEPA Consent\displaymaps.dll” (assuming you installed AutoDEPOMOD to 'C:\SEPA Consent' - if not change the path to where it was installed)
    • if successful you should get a an 'OK' successful dialogue.

Because of the macros in the Excel workbooks (FFMT.xls and FFin.xls) there may also be an issue with macro security depending on how your IT support have set up users on your computer.

An alternative approach is to run AutoDEPOMOD on a virtual Windows XP machine within Windows7; refer to Wikipedia for an explanation of this concept.  This has been successfully implemented using VirtualBox , but requires a copy of the Windows XP operating system.