3

Continuity

Resource/reserve estimation depends first and foremost on a geological model that provides a sound, confident
expectation that a well defined volume (deposit/domain) is mineralized throughout. Without this explicit decision
regarding geological continuity of a delimited mineralized zone, neither estimates nor classification of mineral
inventory is possible. (Sinclair and Blackwell, 2000, p. 34).

near certainty with which the best-defined reserve

In Chapter 3, continuity is defined in relation to category should be known (by observation and very
mineral inventory estimation, and the concept of limited interpolation) and the decreasing certainty of
a mineral deposit consisting of several distinct continuity in other categories of resources/reserves.
domains of continuity is presented. Several case During the 1980s, many gold exploration and pro-
histories are discussed to illustrate the distinction ducing companies placed too little attention toward
between geologic continuity and value continu- confirming the physical continuity of mineralization
ity, as well as to review some of the methods prior to an actual production decision (e.g., Clow,
available for studying continuity. Reference is 1991; Knoll, 1989). The resulting errors in estimat-
made to both classic and new approaches to con- ing metal grades and ore tonnages contributed to the
sidering continuity. early closing of several mines and the abrupt termina-
tion of plans for production at others. More reliable
estimates of mineral inventories require better under-
standing of continuity as a prelude to detailed mineral
3.1: INTRODUCTION
deposit appraisal.
Continuity is a topic of international concern in the Two types of continuity are recognized in min-
study of mineral deposits and the classification of min- eral inventory studies (Sinclair and Vallée, 1994),
eral inventories. This characteristic is an important geologic and value continuity. Definitions are sum-
parameter in several national resource/reserve clas- marized in Table 1.4. The following discussion of
sification systems used to describe formally those continuity is adapted largely from Sinclair and Vallée
parts of a mineral deposit that can be regarded as (1994).
being well-defined assets of mining and exploration
companies. Examples of such systems are those of
3.2: GEOLOGIC CONTINUITY
the United States (USGS, 1980), Australia (AIMM,
1988), and Canada (National Policy Statement 2A Geologic continuity is the physical or geometric oc-
of the Canadian Security Administrators). These re- currence of geologic features that control localization
source/reserve classification schemes describe the and disposition of mineralization. These controlling

59
60 APPLIED MINERAL INVENTOR Y ESTIMATION

Low
1 ppm Diamond pipe
Alluvial Sn, Au
1–30 ppm Rustenburg Pt Stratiform Disseminated Vein Au
Au Au

0.1% Porphyry Mo Mary K. Stockwork W

Porphyry King Island W Vein Sn/W
Cu
1% Stratiform Sn
Stratiform Ni Redross Ni Mississippi Valley Pb Zn
Proportion of Ore Mineral Stratiform Pb-Zn
Volcanic Cu
10% (Unmetamorphosed) Pb Zn Au

Broken Hill Pb Zn Andean Pb Zn

Woodlawn 1 lens
(Metamorphosed)

30%
P
Weipa Jamaica
bauxite bauxite

60%
Fe

Coal
High

Evaporite
100%

High Geologic Continuity Low

Figure 3.1: Qualitative relation of geologic continuity as a function of ore mineral abundance. The diagram is useful in showing
the relative difficulties of obtaining mineral inventory estimations in average deposits of the various classes shown. The
concept of geologic continuity is illustrated schematically along the x axis. Redrawn from King et al. (1982).

features can be lithologic or structural, primary or sec- the sample size – in this case, the zone of dissemi-
ondary, and commonly there is a complex interplay of nated mineralization rather than the dispersed mineral
more than one control. Superimposed metamorphic, blebs.
structural, or alteration processes can disrupt (or en- Geologic observations regarding the nature of pri-
hance) an originally continuous body. Geologic con- mary or secondary features is the input from which
tinuity is a geometric feature and a function of scale; the physical continuity of a mineral deposit is inter-
increasing continuity within a mineralized zone can preted. This geologic information is based on some
be imagined (cf. King et al., 1982) in the progression combination of surface observations, drilling, and un-
from widely dispersed mineral grains through larger derground information that provide the basis for ob-
blebs and semimassive ore to massive ore (x axis in serving and recording the main features of the min-
Fig. 3.1). This is a useful if simplistic view because eral concentration of interest (mode of occurrence
the relative scales of sample size and the size of min- and spatial distribution) and the major features con-
eralized blebs also must be taken into account. For trolling mineral distribution: intrusion; volcanic or
example, 10-m blastholes in a porphyry-type deposit sedimentary layer; faults or shear zones; and folds,
are many orders of magnitude larger than the individ- stockwork, and so on. The methods that can be used
ual blebs of ore minerals. Thus, physical continuity and their effectiveness depend on the level of infor-
of mineralized ground should be viewed in terms of mation available and on the geologic framework and
 CONTINUITY 61

geologic attributes are directional or anisotropic in

shoot
nature, and differ in their character as a function of
direction in space. Several examples emphasize the
ore
rock unit
to be hibiting
Volume from Vein-in (shale, etc.) importance of this attribute of anisotropy. Within a
d e d
exclu block zone of sheeted veins, it is evident that the physical
reserve
continuity of a single vein is more extensive within
d
Indicate

Drift the plane of a vein than across the vein. Similarly,

it is common that the regular array of sheeted veins
Figure 3.2: Example of the importance of using geologic in- has greater physical continuity parallel to the plane of
formation to interpret physical continuity of ore. This ideal- the vein than across that plane. A syngenetic massive
ized example shows the vertical projection of a vein. Recog- sulphide deposit generally is more extensive paral-
nition of the local discontinuity in the vein depends on
lel to bedding than across bedding. Similarly, min-
(i) a knowledge of lithologic control on presence or absence
of mineralization, and (ii) detailed geologic mapping and in- eralization in shear zones is generally more elongate
terpretation. Redrawn from Rostad (1986). within the plane of the shearing rather than across
the shear zone. Anisotropy of shapes of mineralized
zones is a common product of the processes that
deposit type present, but have much in common with form such zones and reflects underlying anisotropic
techniques of stratigraphic correlation and include geologic attributes. This concept of anisotropy is fun-
theoretic studies, alteration patterns, chemical pro- damental in the application of geology to obtaining
files across mineralized structures, mineral associa- high-quality resource/reserve estimates. Experience
tion patterns, and so on, all of which also contribute suggests that preferred directions of geologic con-
to the development of an ore deposit model. tinuity commonly are also preferred directions of
Geologic information is used to interpret explic- grade continuity, as illustrated in Fig. 3.3 for the South
itly, systematically, and in three dimensions (cf. Sides, Tail zone of the Equity Silver Mine, central British
1992b) the general geologic environment and general Columbia.
extent and character of mineralized ground. Then fol-
low assumptions (interpolations and extrapolations)
about the presence, extent, and limits of a mineral- N
ized structure or mass in relation to the sample con-
trol sites and the known geology (e.g., Fig. 3.2). These
assumptions are based on an understanding of conti-
nuity derived from a geologic framework known only
within limits. For convenience, deposit types can be
grouped into a few basic categories. For example,
King et al. (1982) propose a useful geometric scheme
0 100 m
as follows: massive and/or disseminated, stratiform
(or planar/tabular), vein systems, surficial (resid-
ual), and alluvial (placer) deposits. These descriptive Figure 3.3: Open-pit limits, 1310 level, South Tail zone,
categories can be further subdivided if necessary. Equity silver deposit, central British Columbia. The dashed
Direct geologic observations and correlations are sup- line separates the deposit into two domains, each char-
plemented by indirect geophysical evidence to assist acterized largely by stockwork mineralization. In the north-
in developing a three-dimensional image of the geol- ern (smaller) domain, the predominant veins strike roughly
easterly; in the southern domain, the predominant vein di-
ogy in and around a mineral deposit.
rection is parallel to the length of the open pit. These dif-
In many cases, a particular geologic character per- ferent directions of strong continuity of veins are illustrated
sists in much the same manner in all directions within schematically by the ellipses (the axes of which are propor-
a domain (i.e., a feature is isotropic). However, most tional to semivariogram ranges). After Giroux et al. (1986).
 62 APPLIED MINERAL INVENTOR Y ESTIMATION

and subparallel to the tabular form of a mineralized

Low

Possible low-grade V zone, are useful for recognizing changes in orientation

fine-grained Au Sn
and disruptions in tabular bodies (Conolly, 1936).

di ore
ult
VSn

ffic
Contoured maps of such variables as fracture den-

M
sity, vein density, and grade (one or more elements)
Ssn
Proportion of Ore Mineral

as well as mineral or metal zoning maps for successive

tim e

levels or vertical sections are other useful procedures

e
es rad
at
G

Ni
(e.g., Fig. 1.17). They are particularly useful for eval-
Pb Zn
PC
U uating continuity of equi-dimensional deposits and
for comparing spatial distributions of various metals.
For example, two metals may have been deposited si-
B
ult

multaneously with the result that they have a similar

di ss
ffic
Le

P No deposits – cannot be spatial distribution (e.g., Fig. 6.12), or they may have
both high proportion and
FeP
highly variable
been deposited at different paragenetic stages of de-
High

C
E position, in which case there is a possibility that their
high Homogeneity Low spatial distributions will differ significantly.
Figure 3.4: Homogeneity of mineralization versus ore min- Geologic features that affect physical continuity
eral abundance. As used here, the term homogeneity is of a mineralized mass can predate, postdate, or be
akin to the concept of grade continuity. Highly homoge- synchronous with the mineralization process; hence,
neous ores are relatively easy to estimate with confidence; a detailed geologic history is essential to sorting out
less homogeneous ores are more difficult to estimate. Re-
all possible complexities that might affect an interpre-
drawn from King et al. (1982). Symbols are as follows: E =
evaporite; C = coal; Fe = bedded iron ore; P = phosphate;
tation of continuity. Preexisting structures can them-
B = bauxite; Pb Zn = stratiform lead–zinc; Ni = nickel; selves be physically continuous, but this does not
SSn = stratiform tin; PC = porphyry copper; VSn = tin guarantee the existence of a continuously mineral-
veins; V = gold, silver veins; U = uranium. The diagram is ized zone. Undetected en echelon structures can cause
highly schematic and exceptions exist. uncertainty in developing models of physical or grade
continuity (e.g., Leitch et al., 1991). The effect of
It is important to realize that adopting a deposit faulting or folding, which potentially disrupts min-
model introduces implicit assumptions about both ge- eralized ground, also must be considered. Clearly, a
ologic continuity and value continuity, as implied in detailed geologic evaluation, with particular attention
Figs. 3.1 and 3.4. For example, the well-established to mineralization control and possible subsequent dis-
physical and grade continuity parallel to bedding con- ruption, contributes to the understanding of physical
trasts markedly with the highly irregular geometric continuity of geologic bodies and is an essential pre-
form and erratic grade distribution characteristic of lude to mineral inventory studies.
many skarn deposits. These model-related assump- Generally, the limiting scale on which one needs
tions, built into early resource/reserve estimates, must to define geologic continuity is the size of the se-
be documented explicitly as work progresses. Once lective mining unit. In the case of value continuity,
deposit delineation has reached a sufficient level of the required scale of knowledge is substantially less
confidence, physical continuity can be studied effec- than the dimensions of the selective mining unit. The
tively through the use of many traditional procedures. question of scale clearly is important for samples used
In tabular deposits – the use of structure contours (e.g., in reserve estimation, if for no other reason than the
Fig. 3.8) and isopach maps (e.g., Fig. 2.15b) for eval- constraints of possible mining methods and the im-
uating trends and physical disruptions to trends – is plications to ore/metal recovery. Composites that are
well established. Similarly, Conolly diagrams, based large relative to the size of original samples (e.g., 3-m
on contoured distances from an arbitrary plane near core samples vs. 12-m composites) have a smoothing
 CONTINUITY 63

effect on original grade values; consequently, a min- certain how representative the grades of the control
eral distribution pattern that is highly irregular, as points are of the intervening ground. Generally, the
based on contouring grades of short samples, might structural and/or lithologic zones that localize or con-
appear much more regularly distributed if based on trol mineralization (i.e., zones of geologic continuity)
much larger composites. are the limits within which value continuity is defined.
It is one thing to have identified the structure(s) con-
trolling mineralization, but another thing to have rea-
3.3: VALUE CONTINUITY
sonable expectation that the structure, or a particular
Value continuity is a measure of the spatial character part of the structure is continuously mineralized (and
of grades, mineral abundances, vein thicknesses, or of ore grade) between control points. Grades normally
some other value or quality (or impurity) measure, are continuous over much shorter distances than the
throughout a specified domain of interest. As an ex- dimensions of the controlling geologic structure.
ample of value continuity, grades are said to be contin- In the past, value continuity was examined subjec-
uous over distances for which they show a recogniz- tively by using such traditional techniques as grade
able degree of similarity. Hence, continuity of grade profiles (Fig. 3.5) and grade–contour maps/sections
is linked closely with the concept of homogeneity of (e.g., Fig. 1.17); both are useful techniques and should
mineralization (Fig. 3.4). Whereas a geologic attribute form part of the data evaluation stage in preparation
is commonly a present or absent feature, value con- for a mineral inventory study. Grade profiles along
tinuity is a question of degree. Mineralization may linear samples (e.g., drill holes, trenches) are use-
extend between control points; the problem is to as- ful because they illustrate the spatial character of

4.5

Successive 5-m composites

4.0

3.5

3.0
Au (g/t )

2.5

2.0

1.5

1.0

0.5

0
0 10 20 30 40 50 60

Meters
Figure 3.5: Example of a grade profile (5-m composite grades) along a drill hole in an epithermal gold deposit. The drill hole is
entirely within mineralized/altered volcanic rocks and illustrates different physical continuity for lower grades versus higher
grades. A low-grade population of grades is continuous over greater distance (on average) than is a high-grade population.
64 APPLIED MINERAL INVENTOR Y ESTIMATION

contiguous grades (or nearly so) and closely spaced 0.6

samples over short to intermediate distances. Obvi-

ously, it is useful where possible to examine grade
profiles for drill holes with different orientations 0.5
22 Azimuth
through a deposit. Contoured values (commonly
widely spaced control points) reflect an implicit as-
0.4
sumption that the variable is continuous between con-
trol points. Hence, plots of contoured grade values,
while instructive, must be viewed critically in the

γ (h)
0.3
112 Azimuth

r
study of value continuity; there must be geologic rea-
son to believe that mineralization is continuous within Vertical
the contoured area. For example, grade contours for a 0.2
bench of a porphyry-type deposit probably incorrectly
depicts that part of a mineralized field cut locally by
barren dykes. 0.1

More recently, value continuity has been stud-

ied by the use of autocorrelation functions such as
semivariograms and correlograms that quantify a sta- 0.0
0 100 200 300
tistical or average continuity in various directions
throughout a deposit or a significant domain within Distance (m)
a deposit. In general, these models show an increas- Figure 3.6: Experimental semivariograms (autocorrelation
ing average disparity between samples as the distance functions) for horizontal and vertical directions for the east
domain of the East zone, Huckleberry porphyry copper de-
between samples increases. For many deposits, such
posit, central British Columbia (after Postolski, 1998). Note
measures level off at a sample spacing referred to as that the ranges (distances at which the experimental semi-
the range (i.e., range of influence of a sample). Ranges variograms level off) differ with direction in the deposit
can be the same in all directions (isotropic continu- (i.e., value continuity is strongly anisotropic).
ity) or can vary with direction (anisotropic continu-
ity). Relative ranges can be used to construct ellipses
that demonstrate variations in continuity in differ- continuities in the various domains indicated. All el-
ent directions and from place to place in a deposit lipse axes are proportional to the ranges of influence
(Figs. 3.6 and 3.7). as determined from autocorrelation functions (in this
These quantitative measures of continuity are built case, ranges of semivariograms), which characterize
on an assumption concerning the physical continuity average value continuity as a function of direction.
of a mineralized body. Commonly, this statistical con- The use of autocorrelation functions as a tool with
tinuity is determined with greatest confidence along which to characterize, compare, and contrast value
the main axis of sampling (e.g., along drill-hole axes). continuity quantitatively from one domain to another
Sampling in the other two dimensions is commonly is evident. Such ellipses are also useful in a relative
much more widely spaced (i.e., distance between sec- sense in depicting changing geologic continuity as a
tions and distance between drill holes along these sec- function of direction in space.
tions is much greater than sample spacing along drill Primary factors that affect the estimation of value
holes; see Fig. 3.5). For these less well-sampled di- continuity in a particular geologic environment are:
rections, a conceptual understanding of continuity is (i) mineral/metal concentrations and (ii) mineral dis-
very dependent on geologic interpretation. In Figs. tribution patterns and controls at various scales.
3.3 and 3.7, the long axes of the ellipses are paral- Sample size (support) interacts with these primary
lel to the principal directions of geologic and value factors.
 CONTINUITY 65

the boundaries in the sections where insufficient

assay information is available. (Pan et al., 1993,
p. 382)

No
Different parts of a single deposit can be distinctive

r th
Hornblende–biotite–
feldspar porphyry geologically, and thus can be characterized by dif-

al
ferent models of physical and statistical continuity

ntr
Ce
Hazelton volcanics
(Vallée and Sinclair, 1993). Consequently, for min-
Cu-mineralization South
zone
eral inventory purposes it may be desirable, even nec-
Dominant direction of
essary, to subdivide a deposit into separate domains,
stockwork development using as a basis the geologic features that control or
0 200
Semivariogram characterize mineralization. Even a simple vein can
model
Meters give way over a short distance to a zone of horsetail
Figure 3.7: Main zone, Huckleberry porphyry Cu–Mo de- veins. Similarly, where conjugate fractures control
posit, central British Columbia. The zone is divided into mineralization, one fracture direction can predomi-
three domains (North, Central, and South), each charac- nate in one part of the deposit and the second fracture
terized by a continuity model illustrated schematically for
direction elsewhere in the deposit (e.g., Figs. 2.1, 2.4,
horizontal directions by an ellipse. The radii of the ellipses
represent ranges of influence for Cu as a function of di- and 3.3). In certain cases, a uniform sampling grid size
rection. A circular pattern indicates isotropic continuity of or orientation may not be appropriate for all domains
grades; ellipses indicate anisotropic continuity of grades. or zones of a deposit. The Kemess South porphyry-
See Postolski and Sinclair (1998a). type copper–gold deposit, described by Copeland and
Rebagliatti (1993), is characterized by five distinct
In certain cases, value continuity is approximately continuity domains with differing geologic charac-
related to the concentrations of the metals/minerals of teristics. These authors strongly emphasize the impor-
interest and the geologic deposit model (cf. Fig. 3.4). tance of geologic control in optimizing continuity as-
In particular, the average local variability of grades is sumptions for mineral inventory purposes. Similarly,
directly proportional to average grade. This generality each of the five distinctive lithologic domains at the
is consistent with the concept of proportional effect, in Golden Sunlight gold deposit has its own character-
which the absolute value of an autocorrelation func- istic autocorrelation model for gold-grade continuity
tion (e.g., the level of average differences between (Sinclair et al., 1983).
samples) varies systematically with mean grade, as In practice, many problems in establishing phys-
discussed in Chapter 8. ical continuity are related to shortcomings of the
geologic information base. For example, basic infor-
mation dealing with the geologic framework and the
3.4: CONTINUITY DOMAINS
actual stratigraphy or structure of the rocks hosting
A geological domain is a spatial entity that a deposit may be missing or very sparse because
represents a well-defined mineralized body. A only limited drill intersections are available. In such
qualified domain for mineral estimation should a case, the geologic model, the deposit (geometric)
contain no, or a minimum amount of, “non- model, the derived continuity assumptions, and the
mineralized” materials. The domain boundaries interpreted grade and tonnages are all vulnerable to
are usually defined on the basis of both assay and large changes as new information is obtained.
geological information. Assays are used to de- Some of the types of domains that can be antic-
termine a cutoff criterion for the boundary, while ipated in a porphyry-type deposit are illustrated in
geological information, such as faults, may as- Figs. 2.16, 2.17, and 3.7. In Fig. 2.17, distinction
sist to verify or refine the boundaries and to infer is made between leached, supergene, and hypogene
66 APPLIED MINERAL INVENTOR Y ESTIMATION

zones whose geologic (and ore) character might be ex- Several geologic characteristics of the No. 3 vein
pected to differ from one rock type to another. In order system influence the methodology used to evaluate
to integrate such information into a mineral inventory continuity, namely, changes in orientation of veins,
estimation it is apparent that the basic geologic char- crosscutting faults, a variable alteration halo that lo-
acteristics must be mapped spatially and examined in cally is brecciated and mineralized, and the en ech-
conjunction with assay information. Even where the elon character of the No. 3 vein system (Leitch
host rock appears uniform in a porphyry environment, et al., 1991). These features were recognized through
different domains might result because of different in- detailed geologic investigations of exploration drill
tensities or directions of predominant structures that core and limited underground workings, and their ef-
control primary mineralization (e.g., Fig. 3.7). fects on estimation procedures are worth considering
briefly.
Most of the 118 exploration, diamond-drill-hole
3.5: CONTINUITY IN MINERAL INVENTORY
intersections indicate clearly defined vein intervals; a
CASE HISTORIES
few anomalously thick vein intersections were found
A multitude of methods have been developed for to include both a vein interval and adjacent miner-
studying continuity, not all of which can be illustrated alized brecciated ground. Where intersected in exist-
here. Three gold-bearing mineral deposits of different ing workings, these brecciated margins were found
geologic types serve to demonstrate some of the use- to have limited lateral extent relative to the more
ful approaches that can be applied to developing an continuous vein. The widely spaced drill data also
understanding of both geologic and value continuity provide some insight as to the limited physical ex-
as a prelude to undertaking a mineral inventory study tent of these breccia bodies, in particular, recogni-
(i.e., the Silver Queen epithermal, polymetallic [Zn, tion that they do not extend between any two adja-
Pb, Ag, Au] vein, the Shasta epithermal Au–Ag vein, cent drill holes. In contrast, the vein structure and its
and the Nickel Plate Au-bearing skarn [cf. Sinclair associated mineralization are evident in all 118 explo-
and Vallée, 1994]). ration drill holes, thus establishing the general conti-
nuity of vein material within the controlling structure.
Precious metal grade profiles were found to define
3.5.1: Silver Queen Deposit
vein thickness where marginal breccias occur (Leitch
The Silver Queen property in central British Columbia et al., 1991). Thus, in several drill holes with abnor-
includes polymetallic, epithermal veins that were mally thick vein intersections from the late 1960s for
mined briefly during 1972–1973. Production ceased which logs and core were not available for reexam-
because of too few headings to provide sufficient mill ination, true vein thicknesses were estimated using
feed and liberation problems that led to very low metal precious metal profiles; the excess thicknesses were
recoveries (W. Cummings, personal communications, attributed to noncontinuous breccia zones.
1990). Production and most exploration were centered Vein continuity was investigated in detail by
on the No. 3 vein system, which strikes northwest- means of a structure contour map (Leitch et al., 1991)
erly, extends for about a kilometer of strike length, that displays the following features (Fig. 3.8):
and dips moderately to the northeast (Leitch et al.,
1990). Thickness is variable, commonly in the range (i) The en echelon character of parts of the vein
of 0.1 to 2.0 m. Two recent independent mineral in- system
ventories of the central segment of the No. 3 vein (ii) A substantial segmenting of the vein due to
(Nowak, 1991) indicate reserves of about 700,000 offsets along cross faults, some with more than
tons averaging 0.08 oz Au/t (2.7 g Au/t), 4.8 oz Ag/t 100 ft (31 m) of apparent horizontal movement
(163 g Ag/t), 0.2 percent Cu, 0.8 percent Pb, and (iii) An abrupt large change in vein strike near the
5.4 percent Zn. south end of the system.
 22
00
0E
'

21
NG-3 2500

50
0E

20
50
21

0E
228 lt

00
218
0' 0' Fau

0E
21 '
80 80
' 20
ek

2280'
23 22 0' '
Cre

80
' 80 218 80 2180' 0 100 200 300 ft
' 21 2280'
0'

'
228 2380'
le

25 2180'

00
00 228 208
' 0' 0' 2280' 2380'

25
Co

'
80 0' 0 50 100 m
2380 23 250
' 2380'
0'
27 228 ' 250 2500'
00
' 0' 0 '
250 00
270
0' 25
00 2700' 27
290
0' '
'
00
27
290 270 2700'
0' 0'
'
2700
0'
290
0N
00
23 '
00
29

0N
00
22
0N
50
22

Figure 3.8: Plan of structure contours on part of the No. 3, polymetallic, epithermal, vein system, Silver Queen Mine, central British Columbia (cf. Leitch
et al., 1991). Features that affect continuity include the en echelon character of veins, splays from veins, curvature of veins, and cross-cutting faults. Redrawn
from Sinclair and Vallée (1994).

67
68 APPLIED MINERAL INVENTOR Y ESTIMATION

Examination of underground workings substanti- this contributes to the disparity in the estimates. The
ates this view of geologic continuity, as initially de- polygonal method assumes the unlikely situation of
veloped from exploration drilling. uniform, local continuity that is perfectly known for
Detailed investigations by Nowak (1991) demon- each polygon of influence surrounding each data point
strate well-defined autocorrelation for vein thickness (drill hole); the geostatistical approach assumes a sta-
and various grade measures (Pb, Zn, Cu, Au, Ag) us- tistical continuity to ore grades represented as an av-
ing both variograms and correlograms (Fig. 3.9). The erage continuity by an autocorrelation function.
resulting geostatistical estimates (ordinary kriging) of
grades and tonnage are comparable to, but slightly
less than, global estimates obtained in an indepen- 3.5.2: JM Zone, Shasta Deposit
dent polygonal study (Nowak, 1991). This is a pat-
tern that commonly exists in such comparative stud- The JM structure at the Shasta Mine, northern British
ies. Of course, the value continuity models assumed Columbia, is a highly altered and silicified zone in
for the two estimation methods are very different, and basalt of the Toodoggone Formation. Gold and silver
values of economic interest are associated with small,
epithermal quartz veins that strike northerly with near
vertical dips, located within the much broader altered
(a) zone. Individual veins extend along strike from a few
1.0
to 15 m; many occur in clusters across the strike to
0.8
( 1- r )

define zones of economic interest, as shown schemat-

0.6 ically in Fig. 3.10. Drilling shows the broad, altered
0.4 zone to range to 100 m in width and commonly
0.2

0.0

0 40 80 120 160 200 240 ft

Cor(h) = 0.25+0.35 × Exp 80 (h)+0.4 × Exp200 (h)

1.2
(b)
1.0

0.8
( 1- r )

0.6

0.4

0.2
0.0
25 m
0 60 120 180 240 300 360 ft
Alteration Limit
Cor(h) = 0.17+0.45 × Exp 50 (h)+0.38 × Exp300(h)
Figure 3.10: Schematic representation of the physical dis-
Figure 3.9: Average value continuity for gold accumulation tribution of high-grade values with short-distance continu-
(a) and vein thickness (b) for the No. 3 vein system, Silver ity, relative to spacing of drill sections, as occurs in the
Queen Mine, central British Columbia (see Fig. 3.8) illus- Shasta deposit. Clusters of high grade, northerly striking
trated by an autocorrelation function. In this example, the quartz veins are in a northerly striking altered zone within
autocorrelation function is a modified correlogram (i.e., a unaltered volcanic rocks. Note that high-grade zones can
correlogram for which all values have been subtracted from occur between drill sections and be unrepresented in data
1 so that the form is that of a semivariogram). In this available for estimation; hence, local estimates by projec-
case, exponential models have been fitted to the data (see tion of known grades are subject to large errors. Redrawn
Chapter 9). Data from Nowak (1991). from Sinclair and Vallée (1994).
 CONTINUITY 69

contains one to three separate vein concentrations (a)

across its width. Individual zones of vein concentra-

tions are commonly from less than 1 m to several
meters in width (Nowak et al., 1991). (b)
The two principal mappable features whose phys-
ical continuities are of interest are as follows:

(i) A variably mineralized, broad alteration zone

(c)
(ii) Confined zones of high concentrations of small
quartz veins or quartz-infilled breccia.

Surface exposures, exploration drill holes, and

underground workings all show that the alteration
(d)
zone is a continuous, crudely tabular zone. This zone
appears to have developed outward from various frac-
ture zones now occupied by quartz veins, concentrated
in less well-defined, vertically dipping lenticular Figure 3.11: Conditional simulations of grades above and
masses. Most individual quartz veins appear to have below cutoff, JM zone, Shasta Mine, Toodoggone Area,
a physical continuity very much less than the 25-m British Columbia. (a, b) Two independent horizontal simu-
spacing of drill sections; even clusters of veins com- lations; (c, d) two independent vertical simulations. Black
is ore grade, stippled is altered rock below ore grade. Area
monly do not extend between two adjoining sections.
of simulation is 500 m × 55 m. White areas are unminer-
The clusters of quartz veins are associated with alized or areas for which insufficient data were available to
the highest precious metal grades; understanding their create a simulation.
continuity is fundamental to forecasting the loca-
tions, tonnages, and grades of ore zones. The phys-
ical occurrence of ore-grade material between rel- prove adequate as an estimation procedure. Never-
atively widely spaced sampling information from theless, in such cases of insufficient data, conditional
exploration drilling provides an added component of simulation is a potent tool to understand and illustrate
uncertainty. the character of grade continuity, and may serve as an
A quantitative model for grade continuity (semi- adequate estimation procedure in some cases.
variogram model) has been constructed from produc- In summary, the geologic continuity of the broad
tion data and known geology to demonstrate, through altered zones can be established with confidence by a
simulation, the nature of grade distribution and the variety of routine geologic mapping procedures that
correlation (continuity) problem (Nowak et al., 1991). provide the necessary information to construct a three-
Figure 3.11 is an example of two such conditional dimensional geometric model. Exploration data, how-
simulations done using GSLIB software (Deutsch ever, are much too widely spaced (drill sections are
and Journel, 1998), and clearly demonstrates the in- about 25 m apart along the structure) to provide
tractable problem of estimating local resources where a confident interpolation of grades. In such cases,
the physical continuity of ore shoots is short com- conditional simulation is shown to be a practical
pared with the spacing of available data. This exam- means of clarifying and documenting the problem of
ple demonstrates how important it is that exploration grade interpolation.
data provide insight into the detailed nature of local
grade continuity so that there is clear appreciation
3.5.3: South Pit, Nickel Plate Mine
of whether interpolation is possible. When interpola-
tion is not possible, more closely spaced information Mineralization at the South Pit, Nickel Plate Mine
might be required, or conditional simulation might (south central British Columbia) occurs as three
70 APPLIED MINERAL INVENTOR Y ESTIMATION

erratically with sulphides. The geometry of mineral-

ized zones can be approximated from assay plans and
sections (Sinclair et al., 1994). Almost all of the 371
5500'
surface exploration drill holes used are vertical and
intersect one or more of the sulphide-bearing zones
Elevation (ft)

Layer
indicated in Fig. 3.12. The presence of sulphides in the
"two" appropriate stratigraphic positions in large clusters of
Layer
drill holes demonstrates the geometric continuity of
Layer
Legend:
"one" "zero" the three sulphide zones in three dimensions. Thick-
Calcite skarn
Diorite sill nesses of these zones are variable, and so interpola-
5000' Limestone
tions between adjacent drill holes are approximate.
Gold-grade profile
- 5000' - 4500' - 4000' Nevertheless, the physical continuity of roughly tab-
ular form for each of the three sulphide-bearing zones
Easting (ft)
is established. Physical continuity of sulphide zones
Figure 3.12: Vertical E–W section (1000 N) through the cen- is disrupted locally by thick dioritic sills (Figs. 3.12
ter of the South Pit, Nickel Plate skarn gold deposit, Hedley,
and 3.13).
British Columbia. Labeled vertical lines are exploration drill
holes projected on the plane of the section. Basal lime- A detailed data evaluation included examination
stone and diorite sills are shown as shaded patterns. Rel- of exploration drill-hole profiles (Fig. 3.12 and 3.14)
ative gold grades in zones of skarn (blank) are indicated and probability graphs of gold grades (Fig. 6.6). Pro-
by bar-graph profiles on drill holes. Ultimate pit and original files showed that individual high-grade gold assays
topography are shown for reference. Redrawn from Sinclair
were invariably flanked by very much lower grades
et al. (1994).
(i.e., high grades have very limited physical continuity
of the order of 3 m or less in a vertical direction along
zones of gold-bearing, massive to disseminated pyr- diamond drill holes). No information is available as
rhotite-rich sulphides in a calcite skarn layer strik- to the extent of high-grade continuity in the plane
ing about N–S and dipping about 30 degrees to the of stratification (between drill holes that are sepa-
west (Figs. 3.12 and 3.13). Gold values are associated rated by about 80 ft). The probability graphs identified

1250'

Layer
"zero"
Layer area
"one"
area
1000'
Northing (ft)

750' Legend:

Blocks above cutoff grade

Diorite dykes

- 5000' - 4750' - 4500' - 4259'

500'

Easting (ft)
Figure 3.13: Plan of the South Pit, 5280 bench, Nickel Plate Mine, southern British Columbia. Dotted zones are mapped
dioritic intrusions that disrupt continuity of ore. Black rectangular areas are blocks estimated to be above cutoff grade (i.e.,
> 0.03 oz Au/t). Redrawn from Sinclair et al. (1994).
 CONTINUITY 71

Layer "Two" Layer "Zero"

5 0.15

0.10
3
Au (oz/t)

Au (oz/t)
2
0.05

0 0.00
332.5 339 350.5 358 367.5 547 554.5 561 568.5 577
335 345.5 354 362.5 551 557.5 564.5 573 581

Sampling Interval (ft) Sampling Interval (ft)

Figure 3.14: Two short-grade profiles from vertical exploration drill data illustrate the different nature of continuity of high
grades (left) and low grades (right). In general, extreme grades are isolated from each other by relatively large intervals of
low grades, whereas low grades are more highly correlated spatially (i.e., low grades are more similar over much longer
distances than are high grades). After Sinclair et al. (1994).

this upper population of gold grades with low conti- intellectual flexibility required by mineral inventory
nuity as values above about 17 g Au/t (0.5 oz/t); coin- estimators in adapting methodology to geology.
cidently, this is the cutting factor used at the minesite For the Silver Queen deposit, the traditional
(Sinclair et al., 1994). Results of this study illustrate method of structure contours is particularly useful in
the dramatic impact of accounting or not accounting defining primary and secondary aspects of en echelon
for the different characters of high- and low-grade vein (geologic) continuity. Autocorrelation functions
gold continuity; exploration-based estimates using the were determined to quantify value continuity mea-
inverse squared distance method overestimated grade sures such as follows:
by about 45 percent, in contrast to a block kriging
approach that considered different continuity models (i) Vein thickness (thus contributing to vein geome-
for high- and low-grade gold subpopulations. When try and tonnage estimates)
the differences in continuity are taken into account, (ii) Metal accumulations (grades × thickness), hence,
resulting reserve estimates based on exploration data grades (Nowak, 1991).
closely match production.
Both widely spaced drill-hole information (70-m
spacing) and more localized, closely spaced informa-
tion from exploratory underground workings (about
3.5.4: Discussion
3-m spacing) were used to develop the autocorrela-
In a discourse such as this, it is not possible to con- tion models. The vein system was treated as a single
sider all aspects of continuity in detail. Each mineral domain by “unfolding” the two segments of different
deposit is, in many of its attributes, different from any orientations to a common plane coincident with the
other. The three mineral deposits used as examples, central part of the vein. Alternatively, the vein system
however, present very different geologic character- could have been considered two separate geologic do-
istics that lead to different approaches to the study mains, separated where there is a sharp change in
of geologic or value continuity. Thus, the examples orientation.
provide an indication of the diversity of approaches In the case of JM zone, the relatively sparse ex-
available, the importance of geologic control, and the ploration drill-core assays were adequate to define
72 APPLIED MINERAL INVENTOR Y ESTIMATION

geologic continuity of a broadly mineralized/altered Generally, a substantial amount of data is required

zone, but were too widely spaced (25 m) to charac- to establish a sound autocorrelation model for a min-
terize grade continuity (Nowak et al., 1991). Closely eral deposit. With foresight and planning, such data
spaced underground information, although localized, can be obtained early in the exploration history of a
permitted the development of an autocorrelation func- deposit, although the absolute amount of data required
tion for grade, providing a basis for undertaking con- cannot be forecast in advance. Once an autocorrela-
ditional simulations of grade. These simulations tion model has been established, both grade estimates
clearly demonstrate the absence of ore continuity be- and grade simulations based on such models can be
tween sections spaced at 25 m, and assist in defin- generated. The simulation example for the JM deposit
ing the type of additional sampling information that illustrates the advantage of simulations in appreciat-
would be necessary to document details of value conti- ing grade continuity early in the exploration history
nuity, and serve as a sound basis for mineral inventory of a deposit. At advanced stages of exploration and
estimates and, eventually, production quality control at the feasibility stage, conditional simulations, being
(Nowak et al., 1991). closely tied to increased quantities of grade informa-
The crudely tabular shapes of mineralized zones tion, improve our local understanding of grade conti-
at the Nickel Plate Mine are well demonstrated by nuity and provide sound planning and quality control
two-dimensional views showing grade profiles as bar information for both mining and milling.
graphs along exploration drill holes (Fig. 3.9 and
Sinclair et al., 1993). Geologic continuity of three
mineralized domains (layers) separated by barren 3.6: PRACTICAL CONSIDERATIONS
ground is evident. Autocorrelation studies done in- All the practical considerations summarized for
dependently in each domain indicate that the same Chapter 2 apply here. Assumptions regarding con-
autocorrelation model for gold can be applied to tinuity must be adapted periodically to a continually
all three layers. Grade data within individual lay- improving knowledge of the geologic characteristics
ers demonstrate the common problem of multiple of a deposit.
populations – in this case, two, each of which must
be examined separately with respect to continuity. 1. The adequacy of mineral inventory estimates de-
How continuity for each data population is estimated pends on a thorough appraisal of two types of
or dealt with depends on the data density and spa- continuity: geologic and value.
tial disposition of each population. At the Nickel 2. Confidence in both geologic continuity and value
Plate Mine the low-grade gold population is suffi- continuity requires a thorough understanding of
ciently abundant to warrant estimation of an auto- the general and detailed geology of a mineral
correlation function (semivariogram). However, the deposit and the surrounding area. Models of both
high-grade population occurs as isolated values or continuity types must be refined and verified as
clusters of very few values (i.e., the high-grade popu- the geologic information base and the sampling
lation is not concentrated in large enough volumes for base for a deposit evolve.
a characteristic autocorrelation function to be deter- 3. Models of geologic and value continuity are en-
mined). Instead, high grades were cut to a threshold hanced by strict quality control procedures for
defined by a probability graph analysis of the total data acquisition, interpretation, and modeling.
assay data and involving partitioning of the cumu- 4. Evaluating continuity requires systematic geo-
lative curve into two subpopulations, high and low logic and sampling check work, perhaps using a
grades. This graphic technique has not been widely formal audit procedure. Many methods are tra-
used in the study of grade continuity, but clearly ditional; increasingly, however, computer-based
has potential for such a purpose and warrants wider data analysis software is becoming essential to
consideration. this work. More attention must be given to the
 CONTINUITY 73

three-dimensional aspects of sampling grids, par- Corvo, Portugal; in Pflug, R., and Harbaugh,
ticularly in situations in which the range of value J. W. (eds.), Computer graphics in geology: three-
continuity is shorter than sample spacing in a par- dimensional modeling of geologic structures and
ticular direction. simulating geologic processes; Lecture Notes in
5. Several established evaluation procedures that Earth Sciences, v. 41, pp. 213–228.
have not been used widely in continuity studies Vallée, M., and D. Cote, 1992, The guide to the eval-
can provide useful insight into an understand- uation of gold deposits: integrating deposit eval-
ing of value continuity. They are: (i) the use of uation and reserve inventory practices; Can. Inst.
probability graphs to identify thresholds between Min. Metall. Bull. v. 85, no. 957, pp. 50–61.
grade categories (subpopulations) with different
continuities (cf. Noble and Ranta, 1982); (ii) the
3.8: EXERCISES
use of autocorrelation studies to quantify value
continuity (regardless of whether geostatistical 1. Contrast the different character of geologic conti-
estimates are to be obtained); and (iii) the use nuity and value continuity in each of the following
of conditional simulation as a means of repre- scenarios:
senting and understanding the nature of value (a) A zone of sheeted veins versus a stockwork
continuity. zone
6. The physical continuity of an ore-controlling (b) Massive Cu–Ni versus network Cu–Ni in ul-
structure and the statistical continuity of ore-grade tramafic rocks
material within that structure are fundamental at- (c) A feeder zone of a volcanogenic massive sul-
tributes of a geologic model that serves as a base phide deposit versus an upper, stratified sul-
for mineral inventory estimation. These are not phide sequence of the same system. The
attributes that can be easily quantified in terms problem can be answered effectively by con-
of risk. However, they are attributes that, on av- structing sketches of the various mineral-
erage, can be considered to be characteristic of a ization styles and superimposing ellipses to
deposit type, because deposit types generally are represent geologic and value continuity in a
characterized by particular attributes. relative manner.

2. Table 3.1 provides U3 O8 data (%) for contiguous

samples from eight drill holes (after Rivoirard,
3.7: SELECTED READING
1987). Construct a grade profile for any three con-
King, H. F., D. W. McMahon, and G. J. Bujor, 1982, A secutive drill holes and comment on the continuity
guide to the understanding of ore reserve estima- of both low and high grades (see also Fig. 7.9).
tion; Australasian Inst. Min. Metall., Supplement Assume that individual values are for 2-m samples
to Proc. No. 281, 21 pp. and that the drill holes are vertical, collared on a
Owens, O., and W. P. Armstrong, 1994, Ore re- flat surface, spaced at 25-m intervals, and num-
serves – the 4 C’s; Can. Inst. Min. Metall. Bull., bered consecutively across a mineralized field of
v. 87, no. 979, pp. 52–54. flatly dipping sandstone beds.
Rowe, R. G., and R. G. Hite, 1982, Applied geology:
the foundation for mine design at Exxon Minerals 3. Figure 2.1 is a geologic plan of the surface pit,
Company’s Crandon deposit; in Erickson, A. E. Endako molybdenum mine, central British
Jr. (ed.), Applied mining geology; AIME, Soc. Columbia (after Kimura et al., 1976). Assuming
Min. Engs., pp. 9–27. this level to be representative of the deposit, com-
Sides, E. J., 1992b, Modelling ofgeologic disconti- ment on geologic and value continuity models for
nuities for reserve estimation purposes at Neves– the deposit.
74 APPLIED MINERAL INVENTOR Y ESTIMATION

Table 3.1 Sample values of simulated U3 08 grades in eight drill holesa

No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8

0.79 0.09 0.10 0.62 1.13 0.08 0.12 0.16

0.19 0.09 0.94 0.52 1.32 0.08 0.12 0.16
0.51 0.09 0.10 0.27 2.13 0.00 1.52 0.16
0.56 0.83 0.53 0.35 2.82 0.08 0.62 0.18
1.26 0.16 0.10 0.28 0.62 0.08 0.12 0.42
1.14 0.09 0.10 0.30 2.35 0.08 0.12 0.16
2.47 0.09 0.97 5.46 19.17 0.06 0.12 0.1
5.86 0.82 0.45 25.47 1.81 0.08 0.12 0.45
26.89 1.14 3.16 0.15 9.06 0.08 0.12 0.16
24.07 6.52 5.41 0.15 10.98 0.08 0.12 0.16
20.59 0.24 50.43 0.15 12.05 0.08 0.12 0.16
10.30 0.09 11.17 0.15 3.66 2.10 0.12 0.16
5.31 0.20 0.23 0.88 6.76 0.98 0.12 0.16
57.94 0.09 0.20 0.99 3.37 3.53 0.12 0.16
26.04 0.09 0.33 0.15 0.23 9.63 0.12 0.16
22.34 1.82 0.10 0.56 1.74 20.33 0.12 0.16
11.52 0.09 0.19 0.53 0.21 12.11 0.12 0.16
42.79 0.09 0.22 4.51 0.17 4.17 0.12 0.16
1.50 18.07 0.20 0.25 2.57 1.25 0.12 2.17
9.89 38.72 1.14 0.15 2.68 0.08 0.12 0.23
2.33 27.98 1.04 0.15 0.92 0.69 0.94 0.16
0.67 3.93 0.10 5.00 1.94 0.08 5.60 0.16
1.48 5.81 0.10 4.54 0.17 0.08 0.82 0.16
0.15 0.65 0.10 1.64 0.17 0.19 1.40 0.16
0.42 0.09 0.10 0.15 0.17 0.08 6.77 0.26
0.82 0.09 0.10 0.15 0.17 0.20 18.26 3.36
1.48 0.09 0.10 0.15 0.17 0.30 11.14 1.43
4.72 0.09 0.10 0.15 0.17 0.56 4.82 5.00
6.57 0.09 0.10 0.15 0.17 0.69 3.98 17.88
3.31 0.09 0.10 0.15 0.17 0.08 1.67 1.79
4.13 1.43 0.25 3.04 0.17 0.08 1.42 1.36
11.31 0.32 0.10 9.57 0.17 0.08 0.23 11.84
12.48 0.09 0.10 6.67 0.17 0.08 1.61 1.73
7.68 5.19 0.10 5.95 0.17 0.08 1.58 0.23
12.17 1.74 0.10 0.96 0.17 0.08 1.96 0.53
0.59 0.09 0.10 5.66 0.17 0.08 3.72 0.16
0.15 1.52 0.57 0.58 0.17 0.08 9.16 0.16
1.04 12.20 0.55 0.15 0.17 0.08 3.09 0.16
1.05 2.19 0.10 0.15 0.17 0.08 0.49 0.16
1.73 1.28 0.10 0.15 0.17 0.08 0.12 0.16
1.98 0.21 0.96 0.15 0.17 0.08 0.12 0.16
3.54 0.09 1.08 0.59 0.17 0.71 0.12 0.16

a
After Rivoirard (1987).
 CONTINUITY 75

4. Consider a 100-m (northerly) by 30-m horizon- (a) Comment on the quality of the sampling plan, as-
tal zone of vertical, sheeted veins striking N–S. suming it to be an early stage evaluation.
Individual veins are 2–4 cm wide, can be traced
for 20–30 m along the strike, and are spaced at (b) The grade estimation problem is two-dimensional
10–15 cm intervals across strike. Three lines of at this stage. How would you arrange an array of
samples have been taken across strike: 25, 50, 5 × 10 m2 blocks for estimation? Why?
and 75 m north of the south boundary. Each line
crosses the entire 30-m width of the deposit and (c) Categorize the blocks in your array into several
is composed of six contiguous 5-m samples. groups of relative quality of estimate.