Disposicion de Relaves, El Costo Del Exceso de Agua en La Presa

7/24/2019 Disposicion de Relaves, El Costo Del Exceso de Agua en La Presa

1/6

h e d is p o s a l o f m i n e t a i l i n g s t h e r e a l

c o s t o f e x c e s s w a t e r in r e s id u e s

by J. A . WATES*, P r.E ng., M .S e. (C iv. E ng.),

M.B.A . , M.S.A. I.C.E.

SYNOPSIS

T he p ape r ex am ines the co nseq ue nce s o f da m b uild in g w ith slim e o f hig h w ater co nte nt, an d th ere fo re Iow relativ e

density, such as the tailings usually associated w ith the extraction of uranium . It show s how a Iow relative density

reduces the practical m axim um rate of rise, and hence the establishm ent cost, of slim es dam s built by conventional

m ea ns . F urth er h id de n c os ts o fs uc h ta ilin gs d am s a re e xamin ed .

It is co nclu ded th at de waterin g o f tailin gs b efo re th eir disp osal is th e m ost co st-effectiv e m ean s of inc reasin g th e

rate at w hich a particular dam can be constructed, or of reducing operating or stability problem s on existing dam s.

In the longer term, underground disposal seem s to be a feasible alternative. However, it is pointed out that there

a re oth er fa ctors tha t ca n com pou nd the o peratin g d iffic ulties, an d th at th ese sh ould be inv estiga ted b efo re rem edia l

step s are tak en .

SAMEVATTING

Die referaat ondersoek die gevolge van die bou van dam m e met slyk met n hoe w aterinhoud en gevolglik

n

lae

relatiew e digtheid soos die uitskot w at gew oonlik m et die ekstraksie van uraan geassosieer is. D it toon hoe n lae

relatiew e dig the id d ie p rak tiese m ak sim um sty gtem po en gev olglik d ie vestig ing sko ste v an sly kd amme w at v olgen s

d ie k on ve nsio nele m eto des g ebo u w ord, v erlaag . V erd er w ord d ie v erb orge ko ste v an so da nig e u itsk otdamme o nder-

soek.

D ie gevolgtrekking w ord gem aak dat ontw atering van die uitskot voordat dit w eggedoen w ord, die m ees koste-

e ffe ktiew e m eto de is om d ie tem po w aarteen

n

bep aald e dam g eb ou kan w ord, te ve rh oo g, of d ie bed ry fs- o f stab ili.

teitsprobleme by bestaande damme te verm inder. Cor die langer termyn Iyk ondergrondse wegdoening na

n

uitvoerbare alternatief. D aar w ord egter op gew ys dat daar ander faktore is w at dIe bedryfsproblem e kan vererger

e n dat h ierd ie fak to re o ndersoek m oet w ord vo ord at herstelm aatreels g etre f, o f nu we d am m e ge bo u w ord.

Introduction

It is known, and generally accepted, that slime of low

relative density or high water content creates difficulties

in the building of dams and often leads to instability.

However, the reason for these difficulties is not often

understood.

This paper examines the consequences of dam building

with the slim e of high water content (low relative density)

that is usually associated w ith the extraction of uranium ,

particularly w here flotation plants are utilized to extract

pyrite (and the occluded residual gold) from uranium

ta iling s b efo re th eir disp osal.

Low Relative Density

Various processes for the extraction of uranium are in

use in South Africal. The fundamental difference

betw een the processes lies in the solid-liquid separation

and primary uranium-recovery circuits. The most com-

mon of these are illustrated in Fig. 1. The various

solid-liquid separation m ethods produce different ratios

between solids and water, and hence slime of different

re la tiv e d en sity .

The basic difference in quality between the water bled

from the process after the uranium has been extracted

from the filtrates that is relevant here can be found in the

prim ary extraction process. The Buffiex process, incor-

porating resin columns (continuous ion exchange) and a

sm all solvent-extraction plant, produces a large volum e of

*

Partner, Jones W agener Inc., P.O . Box 1434, Rivonia 2128,

Trans-vaal.

@ The South Mrican Institute of M ining and Metallurgy, 1983.

SA ISSN 0038-223X/S3.00 + 0.00.

solution (barren solution), which is free of entrained or-

ganic matter. The Purlex process, on the other hand,

incorporates only a large solvent-extraction plant and

produces a large volume of water (raffinate) , which

contains entrained organic m aterial.

W here flotation plants are coupled to uranium plants

so that the pyrite, and sometimes the occluded residual

gold, can be extracted, entrained organic m atter adverse-

ly affects the flotation performance. Raffinate (from

Purlex) can therefore not be used as filter wash or in the

repulping of feed to the plant. Barren solution, on the

other hand, can be used for both. Therefore, whereas all

the raffinate produced by the Purlex process must be

discarded, as much as ~ per cent of the barren solution

produced by the Buffiex process can be recycled.

In efficient plants, the raffinate, which is discarded

w ith the flotation-plant tailings, reduces the relative den-

sity to a typical value of about 1,25. Plant inefficiency,

spillage, and gland service water reduce the relative

density further, and in some cases values as low as 1,15

are reached. Compared with relative density values of

1,4, which can be achieved from a Buffiex process or

from conventional gold extraction, this difference is

significant.

The total volume of slurry pumped to the tailings dam

is significantly influenced by the relative density. For

example, about 1 t of water goes to the dam for each ton

of solids in a tailing with a relative density of 1,45, while

about 3 t goes with each ton of solids in a tailing with a

relative density of 1,18. Since the capital, m aintenance,

and running costs of pumps and piping are approxim ate-

ly proportional to the volumE> of slurry pum ped, the COfjt

of disposal per dry ton m illed can be significantly greater

for slim e of low relative density.

JO UR NA L O F TH E S OU TH A FR IC AN IN ST ITU TE O F M IN IN G A ND M ET ALLU RG Y

NO VEM BER DECEMBER 983

25 7


2/6

S O L ID S T O R E S ID U E

O R F L O T T IO N

f i L T R T E

L E C H

F I L T R T I O N

F IL T R T IO N W IT H P U R L E X O R B U F F L E X

S O L ID S T O R E S ID U E

O R F L O T T IO N

L E C H

C C D

f i L T R T E

F I L T R

T

C C O W IT H P U R L E X O R B U F F L E X

Fig. I-C ircuits for solid-liquid separation and prim ary recovery in a uranium plant

C C D

=

co unte rcu rren t d ec an tatio n

C IX

=

C ontinuous ion exchange

SX

=

S ol ve nt e xt ra ct io n

Further to the obvious influence that tailings density

has on pum ping costs, the effect on the establishm ent

cost for the tailings dam can be significant. T he section

that follow s illustrates how the relative density influences

the m axim um practical rate of rise, and hence the cost

of d am s.

R ate o f R ise

In p ractical term s for g old and uranium tailing s, th e

m axim um rate of rise is the rate of d epo sition th at allow s

a lo w en oug h cycle tim e on the d am s d ay p add ock s to

fa cilita te d ry in g o ut d esic ca tio n) o f th e ta ilin gs. D esic ca -

tion of the slim e is essential in the first instan ce to brin g

abo ut co nsolidation a nd the resu lting strength g ain an d,

in th e seco nd instan ce, to crack the slim e and thereb y

reduce th e ratio o f h ori2:on tal to v ertical perm ea bility .

F ro m a p ractical p oint of v iew , slim e m ust be relativ ely

dry to allow access and to fac~ litate h and pack ing of a

25 8

N O V E M B E R /D E C E M B E R 1983 JO U R N A L O F T H E S O U TH A F R IC A N IN S T IT U TE O F M IN IN G A N D M E TA L L U R G Y

}

]

P U R L E X

B U F F L E X

}

P U R L E X

}

B U F F L

re la tiv ely im p erv io us b un d o n th e in ne r, a nd in p artic ula r

on th e o uter, sho uld ers of the d ay paddo ck. In tho se

cases w here m echanical p ackers are em ploy ed, the sur-

face o f the pad dock m ust b e dry en ou gh to p ro vide a

tra ve llin g su rfa ce fo r th e tra cto rs.

T he cycle tim e sh ould also be low enou gh to lim it

rech arg in g o f the ph reatic su rface w ater tab le) w ith in

the slim es d am . E xcessive recharging results in a rise in

the p hreatic surface, an d m ay cau se unstable con ditio ns

at the d am to e if the drain age cap acity o f the artificial

blan ket drain s w hen p resen t, an d of th e natural fou nda-

d ati on m a te ri als , a re e xc ee de d.

T h e m axim u m p ractical rate of rise, and hen ce the

m inim um cy cle tim e requ ired for effectiv e desicca tio n,

stable p hreatic su rface con dition s, an d access req uire-

m ents, is co m m o nly tho ugh t to b e a un iq ue value for a

giv en clim atic region. F or go ld and u ran ium dam s in the

Tran sv aal an d O ran ge Free S tate goldfields, a value o f


3/6

1100

1000

900

800

en

en

70 0

::;:

>-

m

0

60 0

-

>..:

Z

UJ

I-

Z 50 0

0

U

IX :

UJ

I-

.00

~300

20 0

10 0

0

2,5

m per year has been adopted as a maximum rate of

rise2,4. T he argum ent that follow s is intended to illustrate

that the maximum practical rate of rise is not unique,

but that cycle time, which depends on the relative den-

sity of the tailings, should be lim ited.

Effect of Relative Density on Cycle Time

Conventional dam-building technique dictates the

limiting volume of a day paddock. Fig. 2 shows a typical

cross-section through a dam, and illustrates the width

and depth of a day paddock. For a given depth and

width, the length of paddock required to accommodate

8 daylight hours of slime increases as the total volume of

slime increases. Simplified assumptions can be made to

show that the cycle time therefore decreases as the

volume of slime increases. Since the total volume can be

increased by increasing water volume without a corres-

ponding increase in dry tonnage, it follows that cycle

tim e is a function of relative density.

D E P T H Idl

Fig.2-Conventional method of slime deposition with

paddocks

The relationship between water content m ass of water

over mass of solids and cycle time in Fig. 3 was derived

theoretically from the assumption of a typical paddock

depth of 200 mm and from the assumption that the

individual day paddocks are filled in a discrete manner.

For example, if the settled slime would occupy 20 per

cent of the total volume of slurry, then after the super-

natant water had been decanted, the paddock would have

risen by 40 mm. For an annual rate of rise of 2,5 m , the

paddock would have to be repacked and refilled about

62 times, or every 6 days. Similarly, if the settled slime

occupied 40 per cent, the cycle time would be 12 days.

The problems associated with slim ing at low relative

density were further compounded recently as a result of

a change in technique occasioned partially by labour

costs and shortages. Previously, the slim e was led to sm all

paddocks formed by crosswalls at about 60 m centres via

a furrow cut on the day wall. Crosswalls have now been

abandoned on some dams, and the day wall is divided

into longer units that are often more than 500 m long.

During slim ing, the entire surface of the large paddock

is wet, and this is sustained for a longer period than for

the smaller paddocks since the filling time is longer.

R O 1 3 5

1 0

1 2 1 3

1

C Y C L E T I M E D A Y S

F i g

3-Relationship between cycle time and water content

The consequence of this practice is illustrated in Figs. 4

and 5. Fig. 4 shows a dam with 14 paddocks, and Fig. 5

shows the same dam with only two paddocks. If one of

the small paddocks takes a day to fill, then the cycle time

would be 14 days, leaving 13 days for drying of the slime.

The large paddocks would, for a dam of the same size,

take 7 days each to fill, and the cycle time would be only

7 days, leaving the other 7 days for drying.

If it is accepted for the time being that cycle time,

rather than rate of rise, is unique for given tailings and

given climatic conditions, then the allowable rate of rise

can be calculated as a function of water content or

relative density. The relationship betw een allow able rate

of rise and relative density, which is shown in Fig. 6,

was derived on the assumption that the previously

acceptable rate of 2,5 m per year was fixed from ex-

perience w ith gold tm ilings, i.e. for tailings w ith a relative

density of 1,45. This, in effect, assumes that a cycle time

of about 2 weeks is acceptable and should be maintained

as a constant.

In practice, paddocks are not filled discretely but

rather over a longer period, while supernatant water and

some slim e are decanted from the low side of the paddock.

The argument presented above and the relationship

derived in Fig. 2 are therefore not strictly correct for

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

NO VEM BER DEC EM BER 98 259


4/6

N I G H T

P A D DO C K

3

7

8

DA Y

PADDOCK

Fig 4 Fourteen day paddocks

actual practice. H ow ever since the paddocks cannot be

filled com pletely without loss of freeboard in the night

area the actual practice is a com prom ise betw een the

two alternatives. Current practice com bined w ith Iow

relative density therefore leads to a com bination of a

decrease in cycle tim e and an increase in the period

during which any particular paddock is exposed to w et

ting.

Recent experience of operational and stability prob

lem s on som e dam s has confirm ed the hypothesis that

cycle tim e rather than rate of rise should be considered

to be the controlling variable. Failures and severe

operating difficulties with uranium tailings dam s have

occurred at rates of rise w ell below 2 5 m per year. These

have been attributed to high water content rather than

to other factors com m only accepted. Failures of dam s

and operating difficulties are shown on Fig. 6 by a graph

for the rate of rise at w hich the problem s occurred

against the relative density of the tailings for the dam in

question. These cases can be seen to correlate relatively

w ell w ith the calculated upper lim it.

It is certain that skilful m anagem ent of dam operation

and tim eous rem edial m easures could obviate som e of the

problem s associated w ith slim ing at Iow relative densi

ties. H ow ever it is never certain that a dam will be

operated in the m ost skilful possible w ay throughout its

life.

M echanization is likely to becom e m ore popular ow ing

to labour shortages and costs and drier conditions w ill

th erefo re b e req uired .

If it is believed that skilful operation and m anagem ent

could increase the m axim um practical rate of rise then

a m odification such as the one show n for optim al dam

building procedure in Fig. 6 could be adopted.

Since the risk of foundation instability on dolom ites is

high a reduced m axim um allowable rate of rise has been

proposed3. This m axim um is also show n in Fig. 6.

If it is accepted that in practice Iow relative density

leads to operating and stability problem s then it w ould

be prudent to recognize these difficulties and to design

n ew d am s ac co rd in gly .

26 0

N OV EM BE R/D EC EM BE R 1 983

JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY

DAY

PADDOCK

NIGHT

PADDOCK

2

F ig 5 T w o d ay p ad do ck s

Costs

A n exam ple of the cost of the sam e dam pum ps

piping and w ater return system designed for an opera

tion involving 250 000 dry tons per m onth at relative

density values of 1 25 and 1 45 was prepared from Jan.

1983 rates for the construction and power. The dam w as

assum ed to be located at the sam e elevation as the resi

due disposal pachucas in the plant and at a distance of

5 km from the plant. The velocity of flow in the pipeline

w as taken as 1 7 m /s and the pow er consum ption was

based on operating experience. The life of the schem e

was taken as 30 years and the pum ping and m aintenance

costs w ere first inflated at 12 per cent per annum and

then discounted at 20 per cent to arrive at a present

value of costs.

The tailings dam was assum ed to be located on an

even ground slope of 1 :100 and the underdrainage

provided by the natural soils underlying the dam was

assum ed to be poor. The dam was taken as rectangular

with a length equal to tw ice the w idth and w as divided

into two com partm ents. Provision w as m ade for a lined

return water dam and an unlined storm w ater dam . The

capacity of the return water dam w as taken as four days

of m axim um average m onthly return from the slim es

dam . The storm w ater dam was sized to retain the run off

from a storm w ith a recurrence interval of 1 :100 years

and of 24 hours duration on the slim es dam and im m e

d ia te c atc hm en t.

Poor natural underdrainage w as assum ed to necessitate

substantial artificial underdrainage. The dam foundation

was taken to be relatively stable w ith no other geotech

nical problem s. N o special features other than statutory

requirem ents pertaining to pollution control w ere taken

in to a cc ou nt.

Fig. 7 which illustrates the relative costs of tailings

dam s of Iow relative density and of high relative density

shows that there can be a significant increase in the cost

of capital expenditure and pum ping as the relative den

sity decreases. The increased cost is thought to be signifi

cant enough to w arrant an exam ination of alternative

approaches.


5/6

R 9300QOO_-

~ \

T O T A L C O S T IN C L . P R E S E N T

\

~ P U M P I N G

R

7500000

\

'

1 '\

'

T O T A L C A P IT A L C O S T

,;X

-

-...

-

R5500~

~

-- -

R4750000

--..

,

-

4200 000

COST

E X C L U D I N G

----/

PIPELINES

-...

.............

R2600000-----

0 D A T A P O I N T F A I L U R E I N V E S T IG A T E D

. .

+

P O O R U N D E R D R A I N A G E ( D E F IN IT E C A U S E I

N O R IN G F EE D

'

~

g i i

- . . . .

< I

o z

0 :0

: ' :~

: : > U J

: ' :>

~~

: ' :u

3 ,0

/ /

y /

/

/

/

/

a :

< I

U J

> -

a :

U J

a .

< J ' )

w

Ir

t-

U J

: : : < : - 2 , 0

W

< J ' )

1 :

Disposicion de Relaves, El Costo Del Exceso de Agua en La Presa

Documents

Transcript of Disposicion de Relaves, El Costo Del Exceso de Agua en La Presa