1.0 INTRODUCTION

Kilborn Engineering (B.C.) Ltd – (Kilborn) was commissioned by CANMET to
organize a Workshop on the modelling of reactive tailings. The Workshop was
held in Haliburton, Ontario, May 23 – 24, 1990, under the chairmanship of Dr.
Henry Steger of CANMET. The Workshop was sponsored by the MEND
Prediction Committee.

The diversified participation consisted of representatives of CANMET, delegates
from the United States, Sweden and Australia and representatives of the mining
industry, the consulting community, universities and various government
departments. The Workshop roster is presented in the Introduction Section of the
Main Report.

The presentations and discussions concentrated on a simulation model studied and
developed under the direction of CANMET and capable of calculating the
generation of sulphuric acid from reactive tailings. The first, third and eleventh
presentation dealt with this program called Reactive Acid Tailings Assessment
Program, (RATAP.BMT1). The program is an adaptation of the program RATAP
to base metal tailings. The model is not yet available to the public.

Other presentations provided a consulting engineering perspective, the industry
perspective and the point of view of the regulatory agencies with regard to the use
of models. Finally, representatives of the United States, Sweden and Australia
presented their thoughts. Numerous discussions took place related to the use of
programmed models capable of calculating the characteristics of reactive tailings
and their impact on the environment.

The report includes a summary of each presentation and the associated question
and answer period.

2.0 OPENING REMARKS

As chairman, Dr. Henry Steger of CANMET welcomed everyone to the
Workshop. Mr. Keith Ferguson of Environment Canada then indicated that the
Workshop is an important step towards the preparation of a model capable of
simulating the reactive tailings acid flux. The finalization of such a model is more
advanced than is generally known, especially with CANMET’s work on RATAP.

He hoped that the participants would be constructive in their critique of RATAP as
well as discuss other models. He indicated that there is industry resistance against
the use of models to characterize reactive tailings, but the presence at this
Workshop of a large number of highly qualified people would contribute to the
development of a useful model for both industry and government.

3.0 PRESENTATIONS DIRECTLY INVOLVED WITH RATAP

The Reactive Acid Tailings Assessment Program, adapted to base metal tailings,
(RATAP.BMT1), was developed from the model used for sulphidic uranium
tailings, (RATAP). The model calculates the amount of acid caused by oxidization
of the sulphide minerals and, in addition, calculates the concentration of zinc and
copper. In their respective presentations, Messrs. W.J. Snodgrass, K.A. Morin and
B.E. Halbert dealt with RATAP.BMT1.

3.1 Dr. W.J. Snodgrass of Beak Consultants explained that a model must be
capable of producing useful results so that one can make decisions with respect to
reactive tailings. He explained the conceptual content of the model and provided
his opinion on its weaknesses, limitations and future possibilities. Dr. W.J.
Snodgrass spoke about the validation and calibration of the model. He compared
RATAP.BMT1 to other existing models and indicated that RATAP.BMT1
addressed questions more complicated than other models. However because of its
complexity, RATAP.BMT1 is more difficult to use and to understand by the new
or inexperienced operator.

A list of eleven questions (problem statements) used for assessing acidic drainage
and which models can help answer is given on page 2.7. RATAP.BMP can address
Questions 1 through 5:

1. Time to oxidize all pyrite
2. Time of acid generation for tailings area
3. Acid flux from a tailings area
4. Flux of acidity and heavy metals from a tailings area
5. Estimate of uncertainty of flux of above point 4. RATAP.BMP needs some information from Questions 6, 7 and 8 for its use.
6. Water balance
7. Groundwater flow
8. Thickness and soil moisture of unsaturated zone. Other questions can be addressed by other codes:
9. Consolidation (geotechnical based)
10. Equilibrium sequences
11. Control options.

3.2 Dr. K.A. Morin of Morwijk Enterprises Ltd. presented a critique of the
RATAP.BMT1 model based exclusively on the main report because other reports
and the model were not available. He recommended a five step program to
improve the model:

• Halt (government-sponsored) model development for several years.
• Initiate detailed field and laboratory studies for comparison with RATAP
  simulations and for expanding current in-field knowledge of acid drainage
  from tailings impoundment’s.
• Release the model as soon as possible to the public domain, that is, to the
  intended users, to allow an extensive determination of RATAP’s
  acceptability and applicability.
• Several years after release of RATAP, conduct a survey of users to
  determine where RATAP can best be improved to make it more accurate and
  more widely applicable.
• At that point in time, determine whether government-sponsored
  improvements should be undertaken.

During the discussion period, a question related to the possibility of using the
model in its present form was asked. A discussion followed which seemed to
indicate that the model should not be released without modifications, especially
with regard to making it more user-friendly. It is also important to note that the
model RATAP.BMT1 must be calibrated and validated for each specific site. This
is the main reason why the program has not yet been released. It was brought out in
subsequent discussions that calibration of the model is not a model coding change
but is, in fact, incorporating site specific data.

3.3 Mr. B.E. Halbert of Senes Consultants Ltd. ended the presentations devoted to
the program RATAP.BMT1. He gave an overview of the model content, its
calibration and validation at Waite Amulet in Quebec. and at Nordic (Elliot Lake),
and its application at Nordic.

The model consists of five modules which can be verified independently.
Deterministic or probabilistic simulations can be carried out. A good resemblance
was achieved between the predicted values and field observations.

During the discussion period, a question was asked related to the calibration of the
model and if such calibrations had been carried out sufficiently to have confidence
in the utilization of the model and its results. The recommendations summarized by
Dr. K.A. Morin were repeated and the question was asked what to do with
RATAP.BMT1 and how to proceed?

Generally, the consensus of opinion was that RATAP.BMT1 should be made more
user-friendly.

4.0 ENGINEERING AND CONSULTING COMPANIES PERSPECTIVE

In their respective presentations, Messrs. K.D. Phinney, A.V. Bell and A.M.
Robertson related their experience with the use of simulation models.

4.1 Mr. K.D. Phinney of Monenco Consultants Ltd. explained that Monenco acts
as a consulting engineer to various mining companies. Its role is to obtain permits,
to carry out feasibility studies, to design tailings disposal areas, etc.

IF With regard to old sites, Mr. K.D. Phinney indicated that the use of a
model is limited to evaluating the efficiency of various reclamation possibilities
and using field measurements. With regard to new sites, it is important that the
water balance be carefully calculated. Mr. K.D. Phinney gave examples of sites
situated in New Brunswick.

Mr. K.D. Phinney closed his presentation with his remarks related to
RATAP.BMT1. It was his opinion that there are simpler models than
RATAP.BMT1 to calculate the acid flux from sites but felt that RATAP.BMT1
will be useful to study the possible modification of tailing sites chemistry.

During the discussion period, a question was asked related to the conservative
methods used by Mr. K.D. Phinney. Engineers are often forced to use very
conservative methods.

4.2 Mr. A.V. Bell from Nolan, Davis & Associates began his presentation by
indicating that models should be simple tools designed to give answers to specific
questions. Models should be user-friendly. RATAP.BMT1 is a very complicated
model that could form the basis for understanding complicated situations.

He hoped that his presentation would promote some discussion on the application
of RATAP.BMT1 or other models to acid waste rock and further would provide
some input to the modelling community for any requirements they might have for
the calibration or development of models relative to acid waste rock situations. The
characteristics of acid waste rock differ in many ways from those of reactive
tailings, although there are of course similarities relative to the acid generation
chemistry and the impact that acid runoff and leachates have on the environment.

His slide presentation was in connection with the waste rock piles of Heath Steele
situated in New Brunswick. The experimental program involves evaluation of the
performance of different covers and is arranged in four phases. The two first
phases, the identification of the piles to be used for the tests and the installation of
monitoring equipment, are completed. The third phase is underway and consists of
determining the characteristics of potential covers.

He is of the opinion that it will be difficult to develop a model capable of
simulating acid generation in waste rock. Such a model is nevertheless very
important.

4.3 Dr. A.M. Robertson from Steffen Robertson & Kirsten (B.C.) Inc. started his
presentation with his comments on RATAP.BMT1. Dr. A.M. Robertson indicated
that RATAP.BMT1 must be released and tested before we can comment further on
it. The model addresses the reactions very well, but not the flow path. Because we
have to give answers today, we must not wait and must release RATAP.BMT1.
We will do better with the model than without.

The presentation of Dr. A.M. Robertson dealt with the modelling of acid
generation in tailings. He showed slides of the physical conditions which exist in
tailings sites at Westmin and at Uravan. Numerous cracks and other irregularities
are responsible for the entry of oxygen and the change in size distribution from
layer to layer results in randomly located perched water tables. A model will only
be valid if it takes these complex site conditions into account.

Dr. A.M. Robertson also reported on his tests at Faro. A special cover put onto a
waste pile limits the entry of oxygen.

During the discussion period, a question was asked related to the availability of
data which could be used by RATAP.BMT1. Dr. A.M. Robertson indicated that he
has some data and that he had asked some months ago to have access to
RATAP.BMT1.

Dr. Robertson reflected on the storage in waste rock dumps, the storage of acid
salts in the dry state and dissolved state. Predictions in the downstream
environment must take into account this storage and it will control our remediation
measures. A large waste rock dump has a large storage capacity and it can take up
to three years before anything comes out of it. Modelling of acid generation must
be attempted using multiflow channel models which will give a stochastic variation
of the acid generation.

Dr. Robertson described Mines Doyon’s lime consumption. The steep increase in
lime consumption is believed to be caused by flushing, not by generation. Storage
of acid is very important in a waste dump, not so much in a tailings site. So, Mines
Doyon needs to know what the acid consumption will be in 1991, we need these
predictive tools. Not all waste is producing acid at Mines Doyon and our models
must recognize this. Tailings dams have edges, thus the geometry and the materials
properties dictate that the models be broken up into sections.

5.0 REGULATORY AGENCY PERSPECTIVE

Mr. Keith Ferguson presented the perspective of the regulatory agencies with
regard to tailings modelling. He first indicated that regulators are very skeptical
towards models and then made a summary of their needs and the questions they
usually ask. He gave his opinion with regard to the capabilities of RATAP.BMT1
to satisfy the needs and to answer the questions.

Mr. K. Ferguson showed slides of the tailings at Westmin and indicated that
RATAP.BMT1 could find an application at this site.

Mr. K. Ferguson concluded that the regulators have a need for the use of models to
predict water quality and expressed the opinion that the use of RATAP.BMT1 will
provide useful information to the regulators.

During the discussion period, a question was asked related to the validation and
calibration of RATAP.BMT1.

6.0 MINING INDUSTRY PERSPECTIVE

Dr. E. Yanful explained the mining industry perspective with regard to the use of
models to simulate reactive tailings. First, he indicated that the needs of the mining
industry are similar to the needs of the regulators; but in addition, the models must
be able to verify the validity of a certain technology. Mining companies are faced
with the problem of disposal of their mine wastes not only in an environmentally
acceptable manner but also in an economical manner. The main aim of the
companies is to stay in business and to make a profit. Dr. E. Yanful explained that
the principal use of the model is to calculate the generation of acid from reactive
tailings, but the model must also be capable of determining the efficiency of
control options. Dr. E. Yanful concluded that the mining industry has many
modelling needs and that the right model must be used in the right situation.
During the discussion period, a question was asked as to whether the mining
industry considers RATAP.BMT1 useful and capable to satisfy its needs. Dr.
Yanful replied that RATAP.BMT1 must be used because the model is considered
to be very useful. However, he was of the opinion that the model has not been
sufficiently validated and that it must be tested more rigorously.

7.0 PRESENTATIONS FROM FOREIGN DELEGATES

Mr. Bj6rn Sodermark and Ms Marie Collin from Sweden, Dr. Kirk Nordstrom
from the United States and Dr. A.I.M. Ritchie from Australia described
respectively their countries perspective.

7.1 Mr. Bj6rn Sodermark of the Swedish Environmental Protection Agency
explained that work started in Sweden ten years ago on the identification and
classification of reactive tailings deposits. With regard to abatement measures, the
preferred solution was the installation of a cover because this method is long
lasting and requires little maintenance. The next step involved the development of
a model capable of determining how to build an efficient cover. He expressed
doubts if ever there will be a comprehensive program that can be fully trusted to
calculate the final acid flux from reactive tailings for all systems and assumed that
it may not be necessary because each problem is site specific. Mr. Bj6rn
Sodermark then presented the results of a case study in Sweden to determine the
best cover, both from a technical and economical aspect. In Sweden, the
government must provide the extra funding for the installation of covers for top
priority projects. Because of the high cost of the abatement measures, Sweden is
now acquiring a better understanding and developing more complete models.
Thereafter, Ms Marie Collin from Kemakta Konsult AB gave a brief summary of
her work carried out in Sweden with models for the calculation of the effectiveness
of covers. She explained the content of her model and described some examples of
calculations carried out with the model. During the discussion period, Ms. Marie
Collin indicated that it was her opinion that the temperature module incorporated
in RATAP.BMT1 was incorrect. Dr. J.M. Scharer replied that the module has since
been changed completely and takes into account the concerns of Ms Collin.

7.2 Dr. Kirk Nordstrom of the U.S. Geological Survey explained that in the U.S.A.,
one is primarily looking at old sites and that it is estimated that there are some 55
billion tonnes of reactive tailings, most of them from copper and uranium mining.

Is there an acid problem? and must one act and how? are the questions that must
have answers. The two remedial measures considered are source control and
effluent control.

Dr. Kirk Nordstrom gave a brief summary of the investigations and treatments
carried out over the last years. He explained that during the 60’s and 70’s the work
was carried out by federal agencies, research institutes and universities who were
very qualified. In the 80’s, the trend is to engage a consultant and such firms often
do not have the proper expertise.

Dr. Kirk Nordstrom stressed the need for closer collaboration between researchers,
government and industry. He would like to see more use made or more effective
use made of the existing geochemical models, thus taking more geochemistry into
account.

Dr. Kirk Nordstrom gave a brief summary of the geochemical models used in the
United States. He explained the studies carried out at Riverton, Wyoming where
field results were compared to results obtained with geochemical and hydrological
models. He explained the results of investigations carried out at Butte, Montana
and expressed his doubts about one of the conclusions of the study which seemed
to indicate that it would take 20,000 years for significant contaminants to move
down into the groundwater.

He also touched on the work carried out at Whitewood Creek, South Dakota where
the model MINTEQ was used and gave a good resemblance to the measured
values. Dr. Kirk Nordstrom described the investigations being undertaken at Pinal
Creek, Arizona, where a plume of acid from an old storage reservoir poses a threat
to the Phoenix aquifer.

7.3 Dr. A.I.M. Ritchie from the Australian Nuclear Science & Technology
Organization (ANSTO) began his presentation with his comments on
RATAP.BMT1. He expressed his conviction that the model will be an important
tool for regulators and operators to evaluate different control options. Then he
gave, chronologically, a description of the work carried out at ANSTO related to
the production and utilization of models.

The model has progressed from a simple homogeneous model, where the oxygen is
transmitted by diffusion through the pore spaces, to an adjusted model, which takes
into account the transmission of oxygen through diffusion into the reaction zone.
Finally, the model has been corrected taking into account the fact that the oxygen
is also transmitted by convection. He then gave some results of the utilization of
the model.

8.0 FINAL DISCUSSION AND SUMMATION

Mr. K. Ferguson announced that the MEND Prediction Committee has decided to
recommend to CANMET that the program RATAP.BMT1 be released for public
use. It was indeed the consensus of the participants of this Workshop that they
want to use the model. The model should be modified to make it more
user-friendly and the documentation should be improved.

Dr. Henry Steger adjourned the Workshop and expressed his thanks to everybody
for their participation. He thanked Ms Mary Paddon of Kilborn Engineering (B.C.)
Ltd. for her excellent organization of the Workshop. Finally, Dr. Henry Steger
wondered if one should not continue to meet in the future and asked the
participants to think about it.