Denehurst Chemical Safety Ltd


Registrants are encouraged to work together to prepare joint registrations and to facilitate this, there was a pre-registration process in 2008 and companies with interests in the same substances are required to form SIEFs.  Vertebrate animal tests must be shared, but other non-vertebrate and non-animal studies can be repeated – however, experience to date suggests most SIEFs are sharing all relevant data.


Most SIEFs are working well; either as part of a formal Consortium or under a less rigid agreement by a small number of organisations and costs are being evenly shared in a fair, transparent and non-discriminatory manner.


However, there are some limited cases where data holders are attempting to charge what could be considered excessive fees for data access or where consortia or consulting groups are attempting to charge excessive management costs.


Unfortunately, when it comes to money, what is seen as fair by one party is not necessarily considered fair by others and disputes have arisen.  For the 2010 registrations, with many large organisations and existing consortia members making registrations, the process appears to have gone smoothly in most cases. 


For the 2013 deadline and again, no doubt, in 2018, some organisations considered the fees requested for a letter of access to be excessive; REACH is not cheap and there will be high costs, especially where there are few companies to share the costs between.  However, there are cases where the letter of access costs are perhaps too high, often for a variety of reasons considered valid by the lead registrant or consortium.

Agreement on costs for data sharing and letter of access


Options to 'opt-out' in a joint regsitration

Services offered by Denehurst


Denehurst can provide an impartial opion on cost share agreements and data valuation to help as an independant arbitrator.



REACH 2018 registrations  

SIEF help 

Cost sharing methods  


Revised January 2014

Independent valuation

In cases where there is no agreement, independent help may be needed to assess the value of the dossier to be submitted. It is possible that those selling the letter of access have included multiple studies for the same endpoint; for example, five of the registrants have repeat dose toxicity studies and two have full reproduction toxicity tests with a total value of several hundred Euro. However, the 2013 registrant only needs one repeat dose study and one reproduction study and does not need to pay for all of them.


A total data set for a 2013 substance, including repeated dose toxicity studies, but excluding some of the more obscure terrestrial and sediment studies that are infrequently needed, is about € 750 000 and dossier preparation should not normally cost more than €50 000 (including the CSR).  Total costs can rise where consortium fees are levied, or if there are multiple studies


From a survey of disseminated dossiers, there are few dossiers for a 2013 (Annex IX) registration with a value of more than  € 300 000; a SIEF of two should therefore come to about €150 000 each, three, €100 000 and so on and for a SIEF of 30, down to €10 000 each.  Note that large SIEFs typically carry more administration costs.


It is therefore easy to estimate a ‘fair’ value for the dossier share and those requiring a registration should consider this crude estimate before deciding if the fees requested by the Lead Registrant or consortium are reasonable.


The next stage

If the letter of access cost is in the same order of magnitude as the independent estimate, the new registrant is advised to pay the sum requested; if the difference is minimal between perceived value and asking price, the time and trouble spent on haggling and then opting out of the joint registration may not be worth the apparent saving.


If there is a big difference in opinion over value, negotiation is needed and independent advice should be sought. In many cases, differences have been overcome with simple dialogue.  The new registrant must attempt to purchase access to the dossier for a ‘fair’ sum and offers that are clearly inadequate will be rightly rejected by the Lead Registrant. 


If agreement is not reached, the new registrant will need to ‘opt out’.


Opting out

Although the ideal is to be part of a Joint Registration, is is possible to opt out of a joint submission and a registrant can prepare their own dossier using their own data, literature, read-across, modelling etc.  Good reasons are needed to opt out, including ‘disproportionate costs’. Only when it comes to vertebrate animal data must the data holder provide access in a 'fair , transparent and non-discriminatory manner‘.  If even at this stage, the data holder chooses not to provide data, or a cost is not agreed, ECHA are accepting registrations and asking that the parties settle their differences.


Data endpoints not purchased will need to be covered by, for example, existing data, read across, literature surveys, modelling. 


If opting out, a full dossier will need to be prepared and this itself will take time and money; this extra ‘hassle-factor’ must be taken into account when deciding what is an appropriate fee to pay for the letter of access.


The registration stage – and legal issues

A dossier submitted that is not part of a joint registration will draw attention to itself.  Dossiers that have no letter of access for end points where the registrant is not a data holder, and ones with data gaps or waivers that are not fully justified will face more attention.


Therefore, opting out is not a decision to be taken lightly.


Even if the dossier is accepted, data holders who consider that their data has been used without payment will rightly be able to pursue the user of their data through the courts for damages and costs could be awarded against the new registrant. ECHA will leave this to the courts and not play a part in the legal process.


However, if it can be shown that the new registrant (the user of data) had made a ‘reasonable’ offer to the data holder that reflected true commercial value of the dossier or of specific data that had been requested, then the courts will very likely find in favour of the data user.  There is a risk that the data holders will receive nothing, or at best, the offer originally made and face extensive legal fees.


Where the data hold is a consortium or industry body, members must be warned that cases that are lost in this way will ultimately be a cost to all members and due consideration must be made of this risk when setting ‘fair’ letter of access fees. 


Summary - Steps to be taken

Opting out is a big decision and a stepwise approach, keeping options open, is advised.


Phase I - valuation

  1. Assess data value

  2. Consider ‘minimal’ dossier value and a commercial value

  3. Suggest offer to Lead Registrant based on ‘reasonable’ commercial value


Phase II - Opting out of joint registration

  1. Determine data gaps, using all public and read-across options etc

  2. Determine remaining studies to purchase

  3. Propose values for studies to purchase

  4. Make formal offers, perhaps through a legal representative with ultimatum


Phase III -Prepare registration

  1. Prepare new testing for non-vertebrate studies if needed

  2. For registration, prepare CSR / extended SDS

  3. Attempt to obtain 'token', register with your own dossier and explain in the dossier reasons for opting out or why letters of access are not provided


Phase IV- Legal

  1. Prepare to face legal action


Estimating the value of a dossier


Various models are in use and support has been provided by CEFIC and other trade bodies to help industry reach agreement; conflicting ideas are clearly going to play a part. 


Organisations who have historically made major capital investments to assess safety of substances will have a different perspective of ‘fair’ to those companies who have been placing the same substance on the market without data.


The methods given below are based on what is known to have worked in many SIEFs, but will not necessarily apply in every case.  One lesson learnt in 2010 was that these methods need to be agreed at an early stage of SIEF formation.


1. Data you need for tonnage band

A lower tonnage registrant in the 1 - 10 t band need only pay access for the data covered in Annex VII, 10 – 100 t in Annex VIII etc.   It is perhaps fair that as both high and low tonnage registrants need such an endpoint that they both pay the same share of those costs.  Note that only a small proportion of Annex X dossiers have full data sets, and the value of data will be similar for Annex IX and X dossiers; so, 2013 registrants may find the letter of access costs similar to those for > 1000 t.


2. Old for new costs

Old work performed is worth the cost of conducting the study today.  There seems a reasonable agreement among much of industry that this is ‘fair' and indeed research by Professor Fleischer  into costs of testing in Europe has been cited as a guide to value data.


      3. Quality of data

Data quality should be judged on whether it is sufficient to avoid new testing costs.  It is proposed by some that if a report is scored 2 or lower on the Klimisch scale  then it is of lower value; this approach is too simple, but easy to apply. 


4. Surplus data

Only data that is directly relevant to the registrant has value and excessive testing, multiple studies on the same endpoint or to cover specific uses not required by the registrant will have limited value. 


5. ‘Risk factors’

Many consortia have added ‘risk factors’ to study values of up to 30%, stating that the original testing carried a risk and those now on the market wanting to buy data did not take that risk.  To date, this does not appear to have been tested in court, but there does need to be some recognition of past capital investment.



Compensation for management costs

Organising test work is time consuming and although there is no direct correlation between the cost of the study and the effort needed to manage, the relationship is close enough to consider a percentage claim.  A rough estimate would suggest between 5% (studies over €100 000) to 15% (small complex studies < €10 000).  It will vary case-by-case.


Preparing a registration dossier and managing a SIEF is also time consuming and figures in the region of €50 000 seem common for moderately complex dossiers.  Where there are extensive waivers and read-across (reducing direct test costs), these dossier costs may be higher.


Obtaining the ‘token’ from ECHA and administrating the letter of access agreements and funds transfer takes time; an administration fee of €500 - €1000 is reasonable, but avoid getting bogged down in complex legal arrangements.




Worked example:

A substance is to be registered in 2013 by a new registrant where there are already 10 existing registrants from 2010.  It appears that there are a further 10 new 2013 registrants (SIEF members are allowed to communicate and discuss these matters using REACH IT contact details) making a total of 20 by 31 May 2013.


Study endpoints covered by test data (determined by information from the Lead Registrant or looking through the disseminated dossier) include the following:


Note: values are taken from the Fleischer list as typical costs in European laboratories in 2007 and may not necessarily reflect today’s costs for new testing


Complete physico-chemical set                   20 000

Acute oral toxicity                                             1 500

Acute inhalation toxicity                                 11 700

Acute dermal toxicity                                       2 000

Skin irritation                                                    1 600

Eye irritation                                                     1 600

Dermal sensitisation                                       4 000

28 day oral toxicity                                          49 000        

90 day oral toxicity                                        €116 000

1 generation reproduction                           €220 000

Bacterial gene mutation                                   3 200

In-vitro mammalian cell gene mutation         17 000

In-vitro chromosome aberration                    19 200

Waivers / assessments for further toxicity endpoints

Acute fish toxicity, inc analysis                        6 000

Acute Daphnia, inc analysis                            4 000

Acute algal inhibition, inc analysis              €    6 000

Sludge respiration                                            2 200

Ready biodegradation                                     3 900

Waivers for hydrolysis, longer term ecotoxicity and soil/sediment studies


Total laboratory costs                                   €488 900


Total laboratory costs accept in this case that all three acute studies are appropriate and that both 28 day and 90 oral studies are included (the 28 day study can be waived if the 90 day study is performed).  All endpoints needed for Annex IX in 2013.


Administration of laboratory work + 10% =         € 538 000

Dossier preparation, including waivers                  60 000


Total estimated costs, approximate                    € 600 000


Total number of registrants, 20.  Cost each            30 000

Plus letter of access admin etc @ €1000               31 000 total estimate        


These figures also clearly show that if only 3 registrants to share the cost, the share becomes very expensive.

Opting out in a Joint Registration



How to opt-out


First step is to check ECHA guidance on this;


In summary, those who feel that the letter of access costs are unreasonable, need to follow certain procedures.

  1. Ask for letter of access cost breakdown from the Lead Registrant. They are obliged to provide this information. This should include confirmation at least of testing costs and administration costs.
  2. If you fail to obtain a satisfactory breakdown of costs, or feel that the costs are unreasonable, then you need to inform the Lead Registrant.
  3. Make a formal offer to the Lead Registrant based on estimated value for a full dossier.  This canbe  estimated i many cases from the disseminated dossier where it is apparent end-points are covered by new tests, read-across or publications. The approximate number of registrants will also be listed, so the total estimate canbe  divided by the number of registrants to get an approximate individual cost. 
  4. There will be bone-fide administration costs, and figures of €40 000 for a dossier are not uncommon; again, divide this figure by the number of registrants. There should be no significant 'legal' costs; if lead registrants / consortia spent a fortune on legal advice, that is their issue.  It was not necessary.
  5. If the offer is not accepted, then inform your national competent authority and inform ECHA
  6. Your own full dossier (or at least parts covering the endpoints you have opted out of)  must then be prepared using your own data, public data and assessments where necessary. Any animal test data held by the Lead Registrant must be shared on request (if needed) . If failing to reach any agreements on animal data, inform your national CA and ECHA. Remember you cannot perform new animal studies if the data already exist.
  7. All this correspondence must be retained as evidence that attempt have been made to obtain the data.
  8. We recommend that your data sections are sent to the Lead Registrant and ideally, they should merge with their lead-dossier and send as a spontaneous update so you too can be full Joint Registrant. However, this is unlikely.
  9. Your own personal registration must be made; if the Lead Registrants has been legally correct and provided the Joint Registration token (a small administration charge is acceptable, perhaps €500), then submit as a Joint Registrant using options in the IUCLID Dossier Preparation step to confirm you are sending some or all of your own end-points.
  10. If the Lead Registrant refuses to co-operate at all, report to your National Agency and submit as a separate (new) registration.  You will have a different registration number and you will need to pay full fees.
Details of European Commission e mails to report unfair or anti-competitive behaviour are: