Hydrogeochemistry and geothermal characteristics of the White Lake basin, South-central British Columbia, Canada

Hydrogeochemistry and geothermal characteristics of the White Lake basin, South-central British Columbia, Canada

Geothermics 31 (2002) 169–194 www.elsevier.com/locate/geothermics Hydrogeochemistry and geothermal characteristics of the White Lake basin, South-cen...

454KB Sizes 25 Downloads 54 Views

Geothermics 31 (2002) 169–194 www.elsevier.com/locate/geothermics

Hydrogeochemistry and geothermal characteristics of the White Lake basin, South-central British Columbia, Canada Frederick A. Michela, Diana M. Allenb,*, Murray B. Granta a

Department of Earth Sciences, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario, K1S 5B6, Canada b Department of Earth Sciences, Simon Fraser University, 8888 University Drive, Burnaby, B.C., V5A 1S6, Canada Received 27 August 1999; accepted 16 February 2001

Abstract Hydrogeochemistry and geothermal characteristics of the Tertiary White Lake basin are described in order to provide constraints on the hydrogeology and thermal regime of the basin. The basin can be divided into three flow subsystems on the basis of chemical and isotopic variations. The groundwaters evolve chemically from young Ca–Mg–HCO3 type waters in the shallow surficial sediments to Na-dominated waters in the deeper intermediate system. Surface waters and shallow groundwaters collected from wells completed in overburden have undergone extensive evaporation as evidenced by their enriched d18O and d2H composition. Minor evaporation identified in the isotope composition of groundwater from domestic wells completed in bedrock, as well as from springs, suggests a local to intermediate origin for these waters, and perhaps mixing with shallow evaporative waters. In contrast, the uniform isotope signatures of deep basin waters measured both spatially and vertically suggest recharge at higher elevations, and a much deeper circulation system that is essentially isolated from the shallow subsurface. Chemical geothermometry indicates that spring waters and bedrock well waters have equilibrated at temperatures of less than 20 and 60 C, respectively. Groundwaters encountered by deep diamond drill holes, with equilibration temperatures of less than 80 C, are representative of intermediate flow systems, and may serve to modify the heat flow regime in the basin. Regional groundwater flow within the basin is complex due to numerous faults that exert a strong influence on fluid circulation patterns. Transport of heat in the subsurface, which has resulted in variations in the measured thermal gradients across the basin, occurs either at depths greater than those investigated in this study or has been significantly influenced * Corresponding author. Fax: +1-613-520-2569. E-mail addresses: [email protected] (F.A. Michel), [email protected] (D.M. Allen). 0375-6505/02/$22.00 # 2002 CNR. Published by Elsevier Science Ltd. All rights reserved. PII: S0375-6505(01)00009-8


F.A. Michel et al. / Geothermics 31 (2002) 169–194

by the circulation of cooler groundwater in the central part of the basin. # 2002 CNR. Published by Elsevier Science Ltd. All rights reserved. Keywords: Geothermometry; Hydrochemistry; Isotope geochemistry; British Columbia, Canada

1. Introduction British Columbia, Canada is located in an area of recent volcanic and tectonic activity. For this reason, many parts of the Canadian Cordillera have elevated heat flow. For example, there are over 100 known hot or warm springs, and certainly many others that remain concealed by the rugged terrain and dense forest that covers much of the province. Some of the hot springs are known to have volcanic affinities; some are apparently associated with geologically recent tectonism (Cenozoic), and others may have radiogenic sources. Several key areas have been identified as potential high-temperature geothermal resources, but the only site that has undergone extensive investigation is Mount Meager, located north of Vancouver. Low energy costs (oil, natural gas and hydro) have resulted in minimal efforts to further explore and exploit high-temperature geothermal resources in the province. However, with energy costs on the rise, geothermal resource exploration and development will likely accelerate. Intermediate and low-temperature resources invariably overlap high-temperature resources, but may also occur in more diverse geological settings. The relatively high permeability of sedimentary basins coupled with centers of high radiogenic heat production or recent volcanism may be suitable for the development of resources of this type. The White Lake basin (WLB) is located 18 km south of Penticton in the lower Okanagan valley of south-central British Columbia (Fig. 1) and is centered on latitude 49 18’N, longitude 119 38’W. The basin is approximately 400 km2 in area, and contains up to 2400 m of sedimentary fill. The WLB has been identified as having above average heat flow (Jessop and Judge, 1971; Lewis, 1984). An initial geothermal assessment of the WLB showed the area to have an average heat flow of 72 mW m 2 (corrected for topography and glaciation), which, when combined with the low thermal conductivity insulating layers of volcaniclastic rocks near the top of the section, can produce temperatures as high as 90 C at the bottom of the basin (Lewis, 1984). The relatively consistent heat flow across the basin, in combination with measured increases in the geothermal gradients across the basin (35–70 mK m 1 from east to west in Fig. 9), led Lewis (1984) to conclude that both local and regional groundwater flow systems are active in the basin. Lewis (1984) proposed two regional flow models to explain the distribution of heat flow in the basin (i.e. relatively constant heat flow with increasing thermal gradients westward). The first model consisted of water flowing down the near-vertical faults from the hills in the center of the basin and flowing along the bottom of the basin until it can escape at lower elevation to the west and possibly to the east. The model allows for water from the very bottom of the basin to flow upward along the basement contact, carrying with it heat from a greater depth and

F.A. Michel et al. / Geothermics 31 (2002) 169–194


Fig. 1. Location map of the White Lake basin (WLB) in south central British Columbia.

warming the adjacent, shallow rocks. The second model consisted of water flowing in the reverse direction from the hills surrounding the basin but escaping up normal faults to lower elevations in the center of the basin. Lewis (1984) considered the former model to be more representative of the thermal regime based on the fact that hot springs are not observed in the middle of the basin and that it is unlikely that the average heat flux from the crust could be that high. Reconnaissance geochemical studies on the WLB were reported by Michel and Fritz (1982) and Grant and Michel (1983). Clark (personal communication, 1986) undertook a preliminary hydrogeologic investigation of the geothermal potential of the Summerland basin immediately to the north of the WLB. To explain the distribution of heat flow within the WLB, it is necessary to understand and integrate the hydrogeology of the basin with the thermal data. This paper describes the hydrogeological and hydrochemical data so that the thermal data, and thus, the geothermal regime of the basin, can be interpreted. In particular, the hydrochemical interpretations will be used to evaluate Lewis’ proposed regional flow model.

2. Hydrological and geological setting The WLB area is characterized by low mountains, and is bounded to the east by Okanagan and tributary valleys (355 m elevation) and to the west by valleys tributary to the Similkameen drainage system (550 m elevation) (Figs. 2 and 3). The central portion of the WLB consists of a small valley (580 m a.s.l.) completely surrounded by mountains ranging in elevation from 915 to 1525 m a.s.l., some of which


F.A. Michel et al. / Geothermics 31 (2002) 169–194

Fig. 2. Tertiary rocks of the White Lake basin and structural subdivisions (modified from Church, 1979). Cross-section location for Fig. 3 (line A, B, C, D) and the map area for Fig. 4 are shown.

topographically separate the White Lake syncline from the Trout Lake graben to the northwest. Local valleys within the WLB contain a series of small lakes and streams that drain the highland areas into the main Okanagan valley. Although the area is relatively arid, farming occurs in the valleys with the aid of irrigation. On average, annual precipitation totals 283 mm while mean monthly temperatures range from 3 to +20 C at Penticton (Atmospheric Environment Services, Canada). High evaporation losses are evident at White Lake and Mahoney Lake (see Fig. 4) where salt crusts develop on the exposed lake beds throughout the summer. The WLB, also referred to in the literature as the ‘‘Penticton Tertiary outlier’’, is an erosional remnant of early Tertiary (Paleocene and Eocene) volcanic and sedimentary rocks (Church, 1973). The Tertiary rocks in the vicinity of the WLB dip gently to the east and are thickest and structurally lowest near the Okanagan valley where the total thickness reaches 2400 m. The local stratigraphy, as described by Church (1973), is summarized in Table 1.

F.A. Michel et al. / Geothermics 31 (2002) 169–194


Fig. 3. Geological cross-section of the White Lake basin showing the location of the Diamond Drill Holes projected onto the section (modified from Church, 1979) and representative regional flow lines and calculated basement temperatures (Lewis, 1984).

Fig. 4. Location map of sampling sites within the White Lake basin.

The rocks that now form this Tertiary ‘‘basin’’ were once probably a continuous belt of mainly volcanic rocks in central Washington and south-central British Columbia. The pre-Tertiary basement is exposed at several places near the margin of the basin and consists of Triassic or older metasedimentary and metavolcanic rocks.


F.A. Michel et al. / Geothermics 31 (2002) 169–194

Table 1 Tertiary formations of the White Lake basin (after Church, 1973) Formation



Approximate thickness (m)



Fanglomerate with large boulders and blocks of Tertiary and pre-Tertiary rock. Mainly landslide breccia with some intercalated conglomerate and tephrite (augite porphyry)


Lower White Lake

Upper Middle and lower


Mainly light-colored pyroclastic rocks, volcanic breccia, 0–90 with some sedimentary rocks and tephrite (augite porphyry) A thick sequence of interbedded volcaniclastic sandstone 0–1100 and conglomerate with some coal, feldspar porphyry lavas, lahars and pyroclastic rocks.


Not subdivided

Predominantly rhyolite and rhyodacite lava with some pyroclastic rocks and local basal conglomerate



Park Rill Nimpit Lake Kearns Creek Kitley Lake Yellow Lake

Andesite lava Trachyte and trachyandesite lava Mainly pyroxene-rich vesicular basaltic andesite lava Mainly trachyte and trachyandesite lava Mainly anorthoclase-rich lava, augite porphyry lavas and pyroclastic rocks

60–450 120–300 0–120 300 150–550

Mainly boulder conglomerate overlying valley talus with fragments of underlying pre-Tertiary rocks


Springbrook Not subdivided

To the south and west, the pre-Tertiary rocks are extensively intruded by granites, granodiorites and syenites of Cretaceous and Jurassic age (Souther, 1977). The Cretaceous and Tertiary volcanic rocks, confined to the Intermontane Belt and eastern flank of the Coast Plutonic Complex (Fig. 1), erupted at approximately the same time as the evolution and uplift of the Coast Plutonic Complex (Souther, 1977). The Tertiary sequence in the WLB is extensively cut by a complex system of normal faults striking predominantly north–south (Figs. 2 and 3). The region is divided into three structural zones by the Marron fault system, which follows the Marron valley southeasterly to Marron Lake (Church, 1973). At Marron lake, it splits into a weak easterly trending branch, which passes into the Okanagan valley, and a strong southwesterly trending branch, which passes near Twin Lakes and extends into the Similkameen valley (Fig. 2). The western part of the basin, comprising structural zone A, is dominated by a series of strong north–south oriented normal faults with easterly downthrow. Small grabens, such as the Trout Lake graben, occur along the eastern margin of the zone adjacent to the Marron fault system. Typically, the strata here are thin and dip gently to the east. The central and southern parts of the basin, comprising primarily structural zone B, have complex faulting. In general, the strata here are folded to form the White Lake syncline, which is open and plunges gently to the east. The beds are cut by normal faults of widely varying trends that show primarily

F.A. Michel et al. / Geothermics 31 (2002) 169–194


westerly and northerly downthrow. Reverse faults, probably related to concentric folding, are developed where strata are especially thick, such as on the north limb of the White Lake syncline. In the southern portion of structural zone C, the Tertiary succession is thin on the west along the axis of an anticline, and thick near the south end of Skaha Lake, site of the Okanagan syncline. Both folds are open and plunge southeastward. Sediments of Quaternary to Recent age (based on water well records) consist of sandy clay, till, glacial fluvial sands and gravel, and talus. Unconsolidated glacial deposits are restricted to valley floors and range from less than 15 m in tributary valleys to more than 60 m in the vicinity of the Twin Lakes. The Twin Lakes are located within an area of kame deposits, which disrupt the normal northward surface drainage to Trout Lake in the Marron valley. Around White Lake, silt and silty clay layers are underlain by sand, gravel and rock fragments. Unsorted silts, sands and gravels cover the floors of smaller valleys and thin talus layers mantle the upper valley slopes. Bedrock is exposed along bluff faces in a number of areas. Minor changes in slope appear to affect the thickness of the talus and result in small spring discharges at an average elevation of 915 m within the main White Lake valley. Many of the small streams draining the upper slopes disappear into the subsurface where the unconsolidated valley sediments onlap the valley walls.

3. Hydrostratigraphy Potential aquifers in overburden and bedrock have been determined through examination of water well records and Church’s (1973, 1979) geologic descriptions. Overburden materials in the basin generally consist of clay, sand, gravel and bedrock fragments. The sands, gravels and fragmented bedrock have the potential to act as aquifers if hydrological conditions are favorable (i.e. good porosity, permeability, etc.). The thickest accumulations of overburden, and hence the best potential aquifers, occur on the valley floors. The clays and silty sands act as aquitards. Permeability of the bedrock formations is probably provided largely by fractures associated with either fault zones or the tops of lava flows, which cooled so quickly that differential contraction created entire horizons of highly jointed rock (Lewis, 1984). Interbeds of conglomerate and sandstone may also provide sufficient permeability, as may the surficial zone of the weathered crystalline basement rock (Lawson, 1968). However, the more felsic volcanic rocks may tend to weather quickly, producing clays that can seal permeable horizons. Locally, faults are deflected along bedding planes between major units and in some places splay out in passing through thin units. These faults are possible routes for water migration, particularly where the adjacent hanging wall and footwall rocks have been brecciated. Water can migrate up, down, laterally along or across faults depending on the hydrological conditions. The nature of the topography and geology in the White Lake basin tends to create artesian pressure in the flow systems, which will cause groundwater to migrate upward or along faults.


F.A. Michel et al. / Geothermics 31 (2002) 169–194

4. Sampling and analysis Water samples of surface waters, springs, domestic wells and diamond drill holes were collected during four site visits between October 1981 and February 1984 (Fig. 4). Many of the sites were sampled repeatedly to provide data on seasonal variations (Tables 2, 3 and 4). In addition, precipitation samples were collected for isotopic analysis at the Penticton airport during the period of June 1983 to January 1984. Diamond drill holes, DDHs, at sites 2, 3 and 4 (Fig. 4) were drilled in 1978 as part of a uranium exploration program (DDH2, DDH3, DDH4). The P-well at site 1 (DDH1) was drilled as part of a study by Jessop and Judge (1971). All four holes are greater than 300 m in depth. Water samples from DDH3, DDH4 and DDH1 were collected at various depths using a pre-rinsed stainless steel down-hole sampler to determine whether compositional variations are present at different levels. The domestic wells, which have been subdivided according to completion in either bedrock or overburden (Fig. 4), were sampled at household taps; water was allowed to run for several minutes prior to sample collection. Springs were usually located near the valley floor where they were conspicuous by the wet ground and dense vegetation in an otherwise very dry and treeless area. Springs were sampled at the source where possible (Fig. 4). Conductivity, pH and temperature were measured in the field using standard hand-held calibrated field meters. Water samples were analyzed for their chemical (major ion) and isotope compositions in the laboratories of the Department of Earth Sciences at the University of Waterloo. Metals were measured by inductively coupled plasma–mass spectrometer (ICP–MS). Total alkalinity was measured by potentiometric titration with sulphuric acid to a final pH of 4.5. Cl , F and SO24 were measured by ion chromatography. Stable isotope ratios were measured using a gas source mass spectrometer (Drimmie et al., 1991; Drimmie and Heemskerk, 1993). Direct tritium 3H was measured by liquid scintillation counting. Stable isotope data are expressed in the conventional d-permille (d%) notation referenced to Vienna Standard Mean Ocean Water (V-SMOW) for 18O and 2H, the PDB standard for 13C, and the CDT standard for 34S. Analytical errors for 18O, 13C, and 34S are  0.2% and  1.5% for 2H. Direct tritium is reported in tritium units (T.U.) with an uncertainty of  8 T.U.

5. Results 5.1. Isotope geochemistry 5.1.1. 18O and 2H Precipitation samples collected at the Penticton airport, plus one snow sample from the DDH1 site (Fig. 4), and a water sample from a high altitude catchment basin (elevation > 1900 m) in Apex Mountain Provincial Park (ski resort water supply) were analysed to determine a local meteoric water line (LMWL) with which


F.A. Michel et al. / Geothermics 31 (2002) 169–194 Table 2 d18O, d2H and 3H data for White Lake basin waters Site number

Sample name

Diamond Drill Holes DDH1 P-Well, surface

P-Well, 41 m P-Well, 91 m P-Well, 122 m P-Well, 152 m P-Well, 213 m P-Well, 251 m P-Well, 305 m P-Well, 366 m P-Well, 488 m



P-Well, 549 m 78-3 78-4, surface

78-4, 154 m 78-4, 385 m 78-4, 430 m 78-6, surf. 78-6, 61 m 78-6, 91 m 78-6, 122 m 78-6, 152 m 78-6, 183 m

Domestic wells in bedrock 5 Bork, 76 m 6a




Domestic wells in overburden 4 Watts (78-6) 7 Bohn

8 9

McWhinnie Twin Lake Golf Club


Twin Lake Ranch

11a 11b

White Lake Ranch 1 White Lake Ranch 2


d18O (%)

d2H (%)


H (T.U.)

Oct. 81 Sep. 83 Feb. 84 Sep. 83 Feb. 84 Feb. 84 Oct. 81 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Sep. 83 Feb. 84 Feb. 84 Oct. 81 Oct. 81 Sep. 83 Feb. 84 Sep. 83 Sep. 83 Sep. 83 Oct. 81 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Feb. 84

18.4 18.1 18.2 18.1 18.2 17.9 18.1 17.9 17.8 18 18.1 18.1 17.9 17.9 17.9 17 18.4 18.2 17.9 18.4 18.2 18.2 18.2 17.8 18.2 18.2 18.6 18.2

138 139 133 137 132 133 137 132 131 131 128 129 132 127 128 133 135 138 131 140 135 134 139 135 138 138 138 138

b.d.a b.d. b.d. b.d. b.d. – 14 b.d. b.d. b.d. 10 b.d. b.d. – b.d. 21 12 b.d. b.d. b.d. b.d. b.d. 4 b.d. 12 5 – –

Sep. 83 Feb. 84 Sep. 83 Feb. 84 Sep. 83

16 15.7 18.4 18 16.2

133 128 140 139 131

b.d. b.d. 12 b.d. 77

Oct. 81 Oct. 81 May 83 Sep. 83 May 83 Oct. 81 Sep. 83 Oct. 81 May 83 Oct. 81 Oct. 81 May 83 Sep. 83

11.4 7.5 7.7 7.8 13.7 9.1 9.7 14.2 14 16.2 16.2 15.4 15.6

102 93 96 97 119 99 103 117 126 122 126 128 126

67 – – 56 b.d. 8 b.d. 34 48 – 49 – 40

(continued on next page )


F.A. Michel et al. / Geothermics 31 (2002) 169–194

Table 2 (continued) d18O (%)

d2H (%)


Site number

Sample name



White Lake Observatory

Oct. 81 Sep. 83

15.4 15.3

124 130

68 50

Springs 13



St. Andrew Golf Club


White Lake Ranch 1


Twin Lake Ranch

Oct. 81 May 83 Sep. 83 Feb. 84 Oct. 81 May 83 Sep. 83 Feb. 84 Oct. 81 Sep. 83 Sep. 83

16 15.9 16.2 15.8 16.7 16 16.2 16.2 16.3 16.5 12.5

131 131 133 123 130 129 129 127 130 132 116

26 – 26 14 7 – 24 b.d. b.d. 17 33

Surface waters 2 17 18 19

78-3 Marsh Skaha Lake, North End St. Andrew Pond Mahoney L.

b.d. 90 52 98 –

White Lake

9.7 14.8 14.3 (+)0.4 6.5 DRY 13.4 5.3 6 8.3 8.4 7.4 7.2 9.1 8.3 11 11.6 16.2 16.1 15.9 15.2

87 118 121 51 87


Oct. 81 Oct. 81 Feb. 84 Oct. 81 Sep. 83 Oct. 81 May 83 Sep. 83 Sep. 83 Oct. 81 Sep. 83 Oct. 81 Oct. 81 May 83 Sep. 83 Oct. 81 Sep. 83 Oct. 81 Sep. 83 Sep. 83 Feb. 84

118 81 83 94 94 86 86 98 89 104 106 121 126 132 122

– 30 – 53 – 98 98 – – 81 – – 67 – 22

12.7 13.4 9 7.1 13.2 14.2 19 21.3 18.2 19.6 17.2

100 109 84 63 108 109 139 164 138 154 127

26 29 45 28 8 7 32 20 68 b.d. –

21 22

Green Lake Trout Lake

23 24

Taylor Lake Twin Lake 1


Twin Lake 2


Park Rill 1


Park Rill 3

H (T.U.)

Precipitation at Penticton airport

1 Snow > 1860 m a.s.l. a

P-Well Apex Mtn. Resort Water Supply

June 83 July 1 8, 83 July 8 25, 83 Aug. 83 Sept. 83 Oct. 83 Nov. 83 Dec. 83 Jan. 84 Feb. 84 Oct. 19, 81

b.d., below detection. The analytical detection limit for tritium analysis is 8 T.U.


F.A. Michel et al. / Geothermics 31 (2002) 169–194 Table 3 Stable isotope data for carbonate (CO3), sulphate (SO4), methane (CH4) samples Site number

Sample name

d13C in CO3 (% PDB)

d34S in SO4 (% CDT)

DDH1 DDH3 11b 13 15 19

P-Well, surface 78-4, 430 m White L. Ranch 2 Observatory Spring White L. Ranch 2 Mahoney Lake

24.5 8.8 16.5 10.5 14.1 (+)1.1

(+) 19.9

d13C in CH4 (% PDB) 46.7

d2H in CH4 (% V-SMOW) 198

to compare all other isotope data (Table 2). The best fit LMWL, shown in Fig. 5a, is slightly different from the MWL determined for Victoria by Fritz et al. (1987), reflecting the dynamic atmospheric processes occurring as the Pacific air mass moves eastward across the Cordillera. Surface water samples display a wide range of values that generally plot along a trajectory of slopes 3–5 (Gonfiantini, 1986), consistent with evaporation (Fig. 5a). As expected, the most highly evaporated samples were collected in late summer and early fall (e.g. White Lake). Mahoney, Green and White lakes shrink considerably in size during the summer and develop salt crusts on the exposed lake beds. In October 1981, White Lake had completely dried up. Even higher altitude lakes like Taylor Lake (elevation 1460 m a.s.l.) can undergo extensive evaporation. The relation between high conductivity and enriched d18O due to evaporation is illustrated in Fig. 6a. Several of the domestic wells completed in overburden also plot on the evaporation line (Fig. 5a). Since the water would not undergo evaporation while in the subsurface, recharge for these shallow overburden flow systems must be from evaporated surface water in lakes or ponds. Waters from the remainder of the overburden wells exhibit only minor evaporation and plot closer to, but generally still to the right of, the intersection of the evaporation and LMW lines. In general, conductivity and d18O show no correlation in waters collected from overburden wells (Fig. 6a). The spring data, which plot in a fairly tight cluster, also display only minor evaporation (Fig. 5b). The one exception, from site 16, is the result of extensive sprinkler irrigation on fields adjacent to the spring. The remainder of the spring data coincide with data from several domestic wells completed in bedrock, and have an isotopic signature different from any of the surface waters or overburden well waters; even those showing only minimal evaporation. This suggests that the spring waters may be discharging from depth within the bedrock, rather than belonging to shallow flow systems within the unconsolidated sediments. DDH waters display a very uniform isotopic composition throughout the length of the borehole (0 to >500 m depth), and from borehole to borehole throughout the region. The DDH data cluster at an average d18O value of 18.2% (Fig. 5b), which is 2% more depleted than the average value for the spring data, and may represent recharge from a higher elevation. While conductivity for these waters is variable, d18O remain relatively constant (Fig. 6a). This relation may indicate a common


F.A. Michel et al. / Geothermics 31 (2002) 169–194

Table 4 Field measurements of temperature, electrical conductivity and pH for selected White Lake basin waters Date

Temp. ( C)

Cond. (mS/cm)


Diamond Drill Holes DDH1 P-Well, surface P-Well, 91 m P-Well, 152 m P-Well, 213 m P-Well, 251 m P-Well, 305 m P-Well, 366 m P-Well, 488 m P-Well, 549 m DDH3 78-4, surface 78-4, 154 m 78-4, 385 m 78-4, 430 m DDH4 78-6, 61 m 78-6, 91 m 78-6, 122 m 78-6, 152 m 78-6, 183 m

Oct. 81 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Oct. 81 Sep. 83 Sep. 83 Sep. 83 Feb. 84 Feb. 84 Feb. 84 Feb. 84 Feb. 84

10.8 10 11 11.5 11.8 11.8 11 11.3 10.8 15 18 18.2 18.9 11.5 13 15 15 17.5

1,350 1,500 1,510 1,480 1,430 1,600 1,550 1,520 1,470 1,890 2,000 2,130 2,190 750 750 800 1,070 1,100

8.33 8.02 7.98 8.0 8.01 8.05 8.01 8.01 8.1 7.81 7.55 7.75 7.95 8.9 8.81 8.9 8.93 8.97

Domestic wells in bedrock 5 Bork, 76 m 6a Nigilser,140 6b Nigilser,165

Sep. 83 Sep. 83 Sep. 83

12.8 17.5 14.9

785 279 1,200

8.25 9.3 7.9

10.5 9 15 10 10.3 10.1 10.8 9.9 15 13

410 435 550 483 480 355 322 350 383 425

– – 7.2 7.55 7.45 – – 6.8 6.65 –

Site number

Sample name

Domestic wells in overburden 4 Watts (78-6) 7 Bohn 8 9 10 11a 11b

McWhinnie Twin Lake Golf Club Twin Lake Ranch White Lake Ranch 1 White Lake Ranch 2


White Lake Observatory

Oct. 81 Oct. 81 Sep. 83 May 83 Oct. 81 Oct. 81 Oct. 81 Oct. 81 Sep. 83 Oct. 81

Springs 13 14 15 16

Observatory St. Andrew Golf Club White Lake Ranch 1 Twin Lake Ranch

Oct. 81 Oct. 81 Oct. 81 Sep. 83

10.9 9 14.1 17

510 282 195 990

7.65 7.7 – 7.06

Surface waters 2 17 19

78-3 Marsh Skaha Lake North End Mahoney L.


White Lake

Oct. 81 Oct. 81 Oct. 81 May 83 Sep. 83 May 83

8.8 13 12 17.5 20.7 19.5

295 165 135,000 5,800 8,500 2,100

– – 9.7 8.75 8.35 9

(continued on next page )


F.A. Michel et al. / Geothermics 31 (2002) 169–194 Table 4 (continued) Site number

Sample name


Green Lake


Trout Lake

23 24

Taylor Lake Twin Lake 1


Twin Lake 2

26 28 >1860 m a.s.l.

Park Rill 1 Park Rill 3 Apex Mtn.


Temp. ( C)

Cond. (mS/cm)


Sep. 83 Oct. 81 May 83 Sep. 83 Oct. 81 Sep. 83 Oct. 81 Oct. 81 Sep. 83 Oct. 81 Sep. 83 Oct. 81 Oct. 81 Oct. 19, 81

23 12.2 15.5 19.8 9.9 18.8 7.6 10.1 20 9.6 19.5 3.9 8.2 8.8

4,150 1,280 1,930 2,300 388 285 128 320 402 205 254 230 406 45

9 – 9.1 8.95 – 8.5 – – 8.3 – 7.85 – – –

recharge area (elevation) for deep groundwaters, but suggests different pathlengths are required to account for variations in conductivity. Water from DDH2 does not plot in the DDH cluster (Fig. 5b); it lies between the DDH and spring groups (there is no conductivity value for this sample). Due to a blockage within the borehole, the sample collected was only from a depth of 18 m. Lewis (1984) noted that an anomaly in the temperature profile of this borehole occurred within the uppermost 30 m. He attributed this anomaly to ‘‘some surface disturbance’’ that may be either associated with a blockage or a surface thermal disturbance. Nevertheless, if one assumes that the isotopic composition of waters below 30 m is similar to that found in the other DDHs, then the disturbance, and isotopic shift, must be caused by shallow groundwater inflow to the borehole above 30 m. In summary, the stable isotopic data reveal three distinct water groups: 1. deep DDH waters (hole depths >183 m) with an average d18O value of 18.2%; 2. the shallower bedrock derived waters found in domestic bedrock wells and most springs with an average d18O value of 16.2%; and 3. surface waters and water in domestic wells completed in overburden, which display varying degrees of evaporation. 5.1.2. Tritium. In an attempt to distinguish the different groups by age, a number of samples were analysed for their direct tritium contents (Table 2). In general, the tritium content increases with decreasing electrical conductivity (approximate measure of total dissolved solids), suggesting a correlation between waters that are chemically immature and residence time (Fig. 6b).


F.A. Michel et al. / Geothermics 31 (2002) 169–194

Fig. 5. (a). Plot of the d18O and d2H data for precipitation and surface water samples. The equations of the Local Meteoric Water Line (LMWL) and an Evaporation Line are defined on the basis of best fit through relevant data. The Global Meteoric Water Line (GMWL) and the Victoria Meteoric Water Line (VMWL) of Fritz et al. (1987) are shown for comparison. (b). Plot of the d18O and d2H data for springs, domestic wells and diamond drill holes. The GMWL, the LMWL and the evaporation line from Fig. 5a are shown for comparison.

Precipitation samples yielded a range from below detection (< 8 T.U.) to 68 T.U. making these data generally unsuitable for comparison with groundwater samples. Surface waters ranged from 22 to 98 T.U.; the one exception (site 2; 0 T.U.) may be indicative of an area discharging older groundwater. The stable isotope data for this sample confirm that this is groundwater that has not undergone evaporation, as compared to the other highly evaporated surface waters. Waters from the domestic wells completed in overburden, which have stable isotopic compositions similar to the surface waters, range in tritium content from not detected to 68 T.U. This suggests that although many of these waters are relatively young, older partially evaporated waters are also present and could be recharged

F.A. Michel et al. / Geothermics 31 (2002) 169–194


Fig. 6. (a). Log electrical conductivity (mS/cm) versus d18O (%) for basin waters. (b). Electrical conductivity (mS/cm) versus tritium content (T.U.) for basin waters.

from surface water bodies. Alternatively, some of these waters could represent partially evaporated modern precipitation that has recently infiltrated. The spring waters yield a similar age range with tritium contents of not detected to 33 T.U. The bedrock domestic well waters range from not detected to 77 T.U., with those waters isotopically similar to the spring waters yielding the highest tritium values. The sample from DDH2, which may contain a component of shallow groundwater, contains some tritium, while samples from DDH3 are consistently tritium


F.A. Michel et al. / Geothermics 31 (2002) 169–194

free. DDH4 and DDH1 have low to no tritium in most water samples. This suggests that the deeper groundwaters are generally older than 30 years (Clark and Fritz, 1997). 5.1.3. Other isotopes Several samples were also analysed for carbon isotope ratios (Table 3) to examine the source of dissolved carbon in the groundwaters and the origin of methane gas discharging from DDH1. The DDH1 sample clearly indicates the production of carbon dioxide during the oxidation of organic compounds, either immature coal or methane (Clark and Fritz, 1997). The White Lake formation, which is intersected by DDH1, does contain numerous immature coal seams, and thus, coal is the most likely source for the methane and CO2. The addition of biogenic carbon dioxide to the groundwater would also explain the high bicarbonate concentrations found in the DDH1 waters, compared to other deeper DDH waters. The d13C value for DDH3 suggests an atmospheric source for carbon, and a single 34 d S analysis of dissolved SO4 for that well (Table 3) is consistent with marine sulphate or coal of Tertiary age and, therefore, excludes sulphide oxidation as the sulphur source. The three samples from overburden wells and a spring (sites 11b, 13 and 15) yielded d13C values for dissolved carbonate that indicate a mixed source from soil CO2 and carbonate dissolution or atmospheric CO2, and d13C for a surface water sample (site 19) possibly represents fractionation during evaporation. 5.2. Water geochemistry Field measurements of the various water types (Table 4) show that the waters in the White Lake basin are geochemically highly variable. Conductivity values for most surface waters, springs and domestic well waters are generally below 1000 mS/ cm, whereas the highly evaporated lake waters are all over 2000 mS/cm; the highest measured value for Mahoney Lake (Fall 1981) was 135,000 mS/cm. The DDH waters have higher conductivities (750–2200 mS/cm) than the shallower groundwaters, as would be expected due to their longer residence times in the subsurface. The waters from DDH4 have much lower conductivities than the waters from DDH3 or DDH1. This suggests that aquifer composition may have an important control locally on the groundwater chemistry. The chemistry of the groundwaters will be influenced by many factors, including mineralogical composition of the aquifers, length and depth of the flow system, subsurface temperature, age of the water (residence time), and whether or not evaporation has taken place. The chemical data for the various waters are displayed on a Piper diagram in Fig. 7, and selected samples are presented in Table 5. Samples with low total dissolved solids (TDS) contents are Ca–Mg–HCO3 waters. These waters are from springs and overburden wells and represent short, shallow flow systems. As the TDS content increases, the waters in this group evolve and plot further right on the diagram due to proportionally larger increases in Na, SO4 and Cl. Water from the 165 m well at site 6b (Fig. 7), which is completed in bedrock, is part of this group. A longer subsurface residence time and higher TDS content

F.A. Michel et al. / Geothermics 31 (2002) 169–194


Fig. 7. Piper plot showing the chemical variability of selected surface and groundwaters in the White Lake basin. Numbers refer to site numbers in Table 5.

places it further along the path of ‘normal’ chemical evolution. The waters from DDH3 may have evolved along this ‘normal’ path. They have the highest dissolved solids content due to large increases in Na, Cl, and SO4 concentrations while the Ca, Mg and HCO3 concentrations have decreased. Several of the lakes that have undergone significant evaporation, as evidenced by their higher TDS contents and heavy isotope enrichment, were also analysed. Green Lake and Mahoney Lake (sites 21 and 19, respectively) show significant enrichments in Na, K, Mg, Cl and SO4, while Ca is nearly absent; bicarbonate contents remain elevated. In the very highly evaporated Mahoney Lake sample (conductivity= 135,000 mS/cm), K, Mg, and SO4 display exceptionally high concentrations. As a result of the elevated SO4 values these evaporated surface waters plot in a separate region from the shallow groundwaters previously described. However, the evolutionary trend visible due to increasing evaporation parallels the ‘normal’ evolutionary pathway (Fig. 7). The surface waters from White Lake (site 20) and the groundwater of nearby Observatory Spring (site 13) both have undergone cation exchange, where Ca and Mg have been lost and Na has been added. Two samples from Observatory Spring are shown in Fig. 7. The sample plotting between the Ca–Mg–HCO3 shallow groundwater group and the second Observatory Spring sample was collected in February and is considered to be a mix of the two end members. The White Lake waters (site 20) have also evolved through evaporation from waters similar to the


F.A. Michel et al. / Geothermics 31 (2002) 169–194

Table 5 Geochemical data for selected surface and groundwaters of the White Lake basin. Concentrations in mg/l Na






Mn HCO3 Cl



DDH1 P-Well, surface Oct. 81 May 83 41 m Sep. 83 91 m Feb. 84 122 m Feb. 84 152 m Feb. 84 213 m Feb. 84 251 m Feb. 84 305 m Feb. 84 366 m Feb. 84 488 m Feb. 84 549 m Feb. 84 DDH3 78-4, surface Oct. 81 May 83 Feb. 84 154 m Sep. 83 385 m Sep. 83 DDH4 78-6, 61 m Feb. 84 91 m Feb. 84 122 m Feb. 84 152 m Feb. 84 183 m Feb. 84

400 520 540 432 520 503 509 506 498 517 509 475 480 500 483 500 490 139 136 133 179 181

2.8 2.33 3.33 2.72 3.6 2.75 2.33 1.98 2.08 2.04 2.11 3.94 1.82 1.59 1.19 1.75 2.25 0.43 0.57 0.28 0.46 0.39

19.4 16.3 17.5 23.9 23.8 25.5 24.2 25.3 22.6 25.5 24.7 22.8 44 36.4 40.2 34.5 34.1 64 44.4 57.8 72.5 80.4

1.5 1.16 1.38 1.11 1.05 1.04 1.05 1.08 1.1 1.05 1.09 1.04 0.87 0.56 0.51 0.66 0.7 0.19 0.16 0.17 0.18 0.21

2.11 2.37 3.05 2.24 2.47 2.02 2.15 2.24 2.33 2.24 2.29 2.11 11.1 13.7 12.1 13.1 12.9 2.42 2.33 2.2 3.63 3.37

4.27 4.7 5.8 2.44 3.08 17.1 0.91 1.53 13.2 33.7 56.7 31.8 22.7 7.7 1.09 0.54 0.97 1.77 2.23 0.38 0.71 0.43

0.08 0.06 0.08 0.05 0.06 0.1 0.07 0.08 0.08 0.16 0.28 0.11 0.05 0.05 0.03 0.05 0.05 0.03 0.05 0.03 0.03 0.03

1154 1199 1272 1225 1230 1147 1192 1165 1187 1187 1158 1160 244 233 287 266 167 65 47 54 40 29.1

183 161 172 193 190 194 185 200 191 189 194 189 512 501 477 469 513 218 211 237 298 302

4.1 4.92 3.38 4.48 4.56 4.52 4.75 4.93 4.52 4.6 4.48 4.68 5.72 6.3 4.72 4.88 4.35 5.16 5.16 5.76 6.4 6.58

2.77 9.29 8.39 1.32 2.16 24.4 9.1 16.2 14.6 22 35.1 38 394 319 326 292 305 102 102 98.4 164 147

7.5 10.9 12.5 14.3 8.5 12.7 8.5 8.2 8.9

Domestic wells in bedrock 5 Bork, 76 m Sep. 83 Feb. 84 6a Nigilser,140 Sep. 83 Feb. 84 6b 165 Sep. 83

245 244 69 72.3 160

1.6 1.03 0.12 0.11 1.83

2.54 1.83 2.24 2.97 103

1.54 1.27 0.08 0.24 44.1

0.82 0.61 0.15 0.21 5.65

5.08 0.82 0.07 0.25 15.7

0.05 0.03 0.05 0.03 0.21

665 659 122 128 420

9.18 7.98 22.6 23.2 140

4.65 5.68 6.29 7.92 0.29

31.4 18.7 21.1 25.7 47.3

7.3 4.7 13 11.9 31

Domestic wells in overburden 7 Bohn Sep. 8 McWhinnie Sep. 9 Twin L Golf Oct. 11b White L R.2 Oct.

83 83 81 81

55.5 52.5 45 22.5

2.43 2.02 2.85 0.38

65 55.3 78 49

25.6 28.2 37.7 14.1

2.31 2.01 2.52 0.41

2.21 1.55 1.4 0.09

0.2 0.76 0.65 0.08

403 381 381 250

4.01 2.37 2.69 1.13

0.35 0.72 0.4 0.5

123 71.9 120 22.5

11 12

Oct. 81 Sep. 83 Feb. 84 Oct. 81 Oct. 81

125 125 66 21.5 23.5

1.8 2.13 1.35 0.28 0.3

52 49.6 68.2 57 43

12.8 12.5 13 15.6 7.25

3.33 3.91 3.95 0.98 0.26

0.12 0.16 0.11 0.05 0.92

0.05 0.05 0.03 0.05 0.05

488 494 390 256 202

31.2 24.9 29.8 1.96 0.6

1.15 1.08 0.72 0.6 0.5

23.5 30.2 19.8 32.1 13.5

Oct. 81 Sep. 83 May 83 Sep. 83 Sep. 83

4900 1550 520 1000 228

730 198 40.4 83 37

28.5 14.9 21 17.9 6.08

2225 508 20.7 23 280

0.44 0.17 0.07 0.28 0.07 0.05 0.98 0.05 2.82 1.1 0.05 0.05 0.05 0.05

1940 710 820 1605 1224

1121 321 79 168 23.9

b.d. 1.06 2.65 4.03 0.66

19440 1010 432 962 701



Springs 13 Observatory

14 15b

St. And. G.C. White L R2

Surface waters 19 Mahoney L 20

White L


Green L


SiO2 7.5 10 6.2 6.3 6.8 6.8 6.3 6.7 6.5 7.5 7.2

12 6.7

10 17.5 15 5

F.A. Michel et al. / Geothermics 31 (2002) 169–194


Observatory Spring waters. The water samples collected from DDH1, just north of Observatory Spring, also appear to have followed a similar cation exchange process to the Observatory Spring waters. However, it is also possible that the DDH1 waters have evolved from a Na–Cl type water through the addition of large quantities of CO2 (as bicarbonate) during oxidation of organics, as discussed earlier. The samples from the bedrock wells at sites 5 and 6a are also Na–HCO3 waters, and display nearly complete cation exchange. Despite their position on the Piper plot (near the highly evaporated Mahoney Lake samples), the waters from DDH4 are not related chemically to the surface waters. Not only are their cation and anion chemistries very different, but the DDH waters have comparatively low TDS contents and isotopically show no evidence of evaporation. The samples do not appear to have undergone cation exchange like the DDH1 samples, and therefore, are considered to have undergone a normal evolutionary process, similar to DDH3, but with a higher Ca/Na ratio. Whereas DDH3 waters essentially contain only Na, the DDH4 waters also contain 30–35% Ca; all DDH waters are very low in Mg. DDH4 water also have a much lower TDS content than DDH3. Because of these differences, the waters from the various boreholes are considered to belong to separate flow regimes of varying length. The water chemistries have probably evolved differently because of interaction with rocks of varied mineralogy, which range from volcanics to carbonate and clastic sediments. 5.3. Geothermometry The chemical composition of thermal waters can provide valuable information on their origin. Dissolved silica and certain cation ratios in deep waters that have experienced prolonged interaction with rocks are usually controlled by temperaturedependent reactions between minerals and the circulating fluids (e.g. Fournier, 1973). Geothermometers represent the equilibria of these temperature-dependent reactions, and geothermometric analyis can indicate the temperature of the reservoir yielding the deep waters. If a geothermometer is to accurately represent reservoir conditions, not only must equilibrium conditions be reached between the ions and the water, but the water must rise relatively rapidly to the surface without chemical alteration due to processes such as evaporation, precipitation or mixing. Several geothermometers were applied to the waters of the WLB: quartz (Fournier, 1973), Na/K (Fournier, 1979; Fournier and Potter, 1979) and K/Mg (Giggenbach, 1988). Table 6 shows the temperatures calculated by these geothermometers for DDH, bedrock well and spring samples. A geothermometric technique that discriminates between ‘‘immature waters’’ and ‘‘fully equilibrated waters’’ coming from deep geothermal reservoirs was proposed by Giggenbach (1988). This technique offers an automatic indication of the suitability of a given water for the application of ionic solute geothermometers. The technique involves combining the Na/K and K2/Mg geothermometers by means of a Na–K–Mg1/2 triangular plot. Because the Na/K geothermometer readjusts relatively slowly in cool environments that encounter rising warm waters and is not strongly affected by mixing with shallow waters, it generally indicates temperatures of deep


F.A. Michel et al. / Geothermics 31 (2002) 169–194

equilibrium. In contrast, the K/Mg geothermometer is very sensitive to cooling and mixing with shallow waters. Fig. 8 shows an Na–K–Mg1/2 triangular diagram for the WLB samples. Spring samples (sites 13, 14 and 15b) and one bedrock well sample (site 6b) plot adjacent to the Mg1/2 corner, which is typical of ‘‘immature waters’’. The Na/K geothermometer frequently gives erroneous indications for such waters. Bedrock well samples from sites 5 and 6a plot only slightly above the Mg1/2 corner, and geothermometer data suggest equilibration temperatures below 60 C and below 20 C, respectively (Table 6). DDH samples plot relatively close to the full equilibrium curve, but have comparatively low equilibrium temperatures (below 80 C). From the geothermometry results it can be concluded that the waters of the WLB have not equilibrated at high temperatures.

6. Flow systems The isotope and geochemical data clearly illustrate the large diversity of groundwater flow systems within the WLB. The groundwaters (springs, wells and DDHs) sampled during this study can be subdivided into three regimes. 1. Wells completed within the overburden sediments that infill the valley floors produce Ca–Mg–HCO3 water that has undergone varying degrees of evaporation.

Fig. 8. Giggenbach Na–K–Mg1/2 triangular diagram for DDH, bedrock well and spring samples in the White Lake basin.


F.A. Michel et al. / Geothermics 31 (2002) 169–194 Table 6 Results of chemical geothermometry for White Lake basin waters Site number (year)

Depth (m)

Quartz (Fa)a

Na/K (FP)a

Na/K (Fb)a

K/Mg (G)a

Na–K–Ca (Fb)a

DDH1 (1984)

41 91 122 152 213 251 305 366 488 549 154 385 61 91 122 152 183 76 76 140 140 165 Spring Spring Spring Spring Spring

39 24 25 27 27 25 27 26 30 29 47 51 34 47 34 33 35 29 17 48 45 81

62 63 67 57 50 44 47 45 46 75 40 50 36 47 23 29 24 64 47 17 13 90 101 110 120 96 95

56 57 61 52 45 40 42 40 42 68 36 45 32 42 19 26 20 59 42 14 10 82 93 101 110 88 87

33 33 30 33 34 36 36 35 36 29 33 31 34 30 37 33 36 41 44 38 50 76 62 60 66 87 76

80 69 79 69 65 59 63 60 62 82 50 57

DDH3 (1983) DDH4 (1984)

#5 (1983) 5 (1984) 6a (1983) (1984) 6b (1983) 13 (1981) (1983) (1984) 14 (1981) #15b (1981) a

45 27

42 41 13 8 23 32 37 16

(Fa) Fournier, 1973; (Fb) Fournier, 1979; (FP) Fournier and Potter, 1979; (G) Giggenbach, 1988.

These waters represent short, local flow systems that occur entirely within the unconsolidated valley sediments. The shallow groundwaters also have high tritium values consistent with their young age. 2. Shallow bedrock wells and springs yield waters that have undergone only minor evaporation. These waters range from Ca–Mg–HCO3 in composition to Na-mixed anion waters, which have partially evolved by the addition of Na and Cl, or by Ca– Na cation exchange. This chemical evolution indicates that the waters have begun to interact with the bedrock encountered along the somewhat longer, but still local flow paths. Tritium values are lower than those for shallow overburden wells (Fig. 6). Geothermometry results suggest that these waters have equilibrated at low temperatures. 3. The older, low-tritium DDH waters display a very uniform oxygen isotope composition but widely differing chemical compositions, suggesting a deeper, intermediate flow system. Of the three drill holes sampled, DDH1 groundwaters displayed a composition indicative of significant cation exchange, with HCO3 as the dominant anion, DDH3 may have either followed a ‘normal’ evolutionary trend to


F.A. Michel et al. / Geothermics 31 (2002) 169–194

become Na and Cl dominated, or alternatively, may have undergone cation change prior to addition of Cl. DDH4 waters are generally similar to DDH3 waters but have a higher Ca/Na ratio. Geothermometry results suggest that these waters equilibrated at temperatures below 80 C.

7. Discussion The model proposed by Lewis (1984) to explain the uniform heat flow measured in the basin as well as the approximately linear increase in thermal gradient from the east to the west side of the basin (Fig. 9) involved flow up dip, from the central part of the basin toward the margins (Fig. 3). This model was favoured by Lewis over an alternative model where water flows from the hills surrounding the basin but escapes up normal faults to the lower elevations in the center of the basin. Both of these models could produce the geothermal gradient and heat flow trend (Fig. 9), but neither one is entirely consistent with the observed geochemical data and hydrogeological system. In Lewis’s first model, recharging groundwater is heated up as it descends along near-vertical faults in the hills in the central part of the basin. The heated waters flow up the basement contact warming the adjacent rocks and result in a progessive increase in the thermal gradients westward. If we interpret Lewis’ model in the broader sense to include both intermediate and regional flow, then the flow lines for

Fig. 9. Geothermal gradients, and measured and corrected (for topography) heat flow values for Diamond Drill Holes (DDHs) shown as projections on the cross-section in Fig. 3.

F.A. Michel et al. / Geothermics 31 (2002) 169–194


the respective systems could be represented as shown in Fig. 3. The flow path for water recharging in the middle of the valley would be largely of intermediate length such that groundwater would discharge in the valleys of the Trout Lake graben and the White Lake syncline. The DDHs have only intersected the uppermost units within the basin (note the projected position of DDH3 appears to be deep in the section). Thus, from a hydrogeological perspective, these flow systems are probably of intermediate path length and are still controlled largely by the local topography. Not shown in Fig. 3 is the possible role of the faults in the Trout Lake graben, which may serve as vertical conduits for the upward flow of deep waters. Our interpretation of the intermediate flow system is consistent with the geochemical and isotopic results discussed in the previous section, as well as with the temperature plots that were presented by Lewis (1984). The plots for the boreholes in the central part of the basin (DDH3 and DDH4) show evidence of significant upward fluid flow, and thus can explain the relatively uniform chemistry of borehole waters. For example, DDH4 was shown to have significant water flow up the borehole (Lewis, 1984), and all data plot in a similar position on the Piper diagram (Fig. 7), with a tendency toward increasing conductivity with depth. This suggests that there may be fresher water inflow at shallow depths within the borehole. Geothermometry results also lend support to the representation of intermediate flow because waters sampled in DDH3 and DDH4 show no strong evidence of a high temperature history. Na/K geothermometers indicate equilibration for DDH waters below 50 C. These results imply a shallower flow path or else significant mixing of deep geothermal waters with colder shallow waters. Lewis interpreted the disturbances in the temperature plots to be caused by warm water entering at discrete depths (likely through fractures) and exiting at higher elevation. Geothermometry results for DDH1, situated to the east of the recharge area, indicate higher equilibration temperatures (although still below 80 C) and may represent greater mixing with deep waters rising up the fault zones of the White Lake syncline (Fig. 3). The interpretation of regional flow, however, is not as straightforward. To provide a driving force for regional flow up dip as proposed by Lewis, a strong downward vertical hydraulic gradient is required from the high elevation hill in the middle of the valley (see Fig. 3). But, given the topographic constraints, it is unlikely that there would be sufficient hydraulic gradient to drive groundwater up dip. In addition, in the western part of the basin, where fluid flow would be directed upward according to the model, the temperature plots show no major disturbances. While the lack of fluid flow does not support or contradict the model, it does suggest that there is an insufficient hydraulic gradient between the fractures intersected at depth to cause fluid flow within the borehole. From a regional flow perspective, there does not appear to be a clear pattern in the chemical data that could aid in the interpretation of the deeper heat flow regime. The relatively uniform isotope composition suggests that the waters from all three DDHs were similar at the time of recharge and may have had a common recharge area or elevation of recharge. However, the waters sampled in each borehole have followed different flow paths since recharge, and have been chemically modified according to the mineralogy of the bedrock units encountered along their flow


F.A. Michel et al. / Geothermics 31 (2002) 169–194

paths. There appears to be an evolution from east (DDH1) to west (DDH3) in terms of the increase in chloride content. The electrical conductivity values (Table 4) indicate that the waters from DDH4 are probably the least evolved, while DDH3 waters have the highest conductivities and the highest Na and Cl concentrations. DDH4 clearly does not follow this evolutionary trend (see Fig. 7), but its chemistry may be more highly influenced by shallow waters (lower conductivities in Table 4). DDH1 waters may be significantly influenced by the package of sedimentary rocks present in the central part of the basin, as suggested by the coal signature seen in the carbon isotopes of bicarbonate and methane gas from this well. Isotopically, the DDH data indicate a similar origin for all deep groundwaters. On the basis of the topography and geology, we speculate that recharge for the regional systems would be at much higher elevations to the west (e.g. Apex Mountain, >1900 m a.s.l.) and that discharge may occur within the main Okanagan valley. Geochemically, these waters would be NaCl dominated, similar to the waters sampled from DDH3. The alternative model presented by Lewis, whereby groundwater flowing within a regional system recharges in the hills surrounding the basin and escapes up normal faults to the lower elevations in the center of the basin, appears to be somewhat more consistent with hydrogeological and hydrogeochemical data. However, if this second model is to be considered, then fluids escaping up normal faults from depth in the center of the basin must show evidence of a geothermal history. The DDHs are relatively shallow throughout the basin, so it is difficult to determine the precise thermal history of their waters. However, there appears to be significant mixing with shallow circulating groundwaters within the intermediate flow system. Mixing may explain why no thermal springs are present in the central valleys. The intermediate and regional flow systems will be controlled by the stratigraphy within and adjacent to the basin, and by the major fault structures that dissect the basin, in addition to topography. The sedimentary units of the Springbrook and Marama formations are discontinuous and of local extent, while the thick sequence of lavas forming the Marron formation are more continuous, despite being offset by faulting. The largescale regional fault zones, like the Trout Lake graben that contains Yellow and Trout Lakes, extend into the basement pre-Tertiary greenstones and schists. Thus, they would present obstacles for regional flow within the basin, as they run perpendicular to the flow directions suggested by Lewis (1984), and would most likely direct water along the various faults within the structure rather than aiding flow across the structure. In the vicinity of White Lake, the fault network (Figs. 2 and 3) could act as a conduit to the main Okanagan valley for waters recharging in the surrounding hills and migrating towards the valley. In view of the complexity of the basin and the underlying structure, it is likely that regional geothermal waters circulate largely in isolation from the upper to intermediate basin flow systems, and are directed by major fault zones. A portion of these deep waters, however, may rise up small faults situated in the Trout Lake graben, and lend a mixed geothermal signature to the DDH waters. Thus, any future efforts to investigate intermediate to low enthalpy geothermal resources in the WLB should include a detailed examination of the fault network.

F.A. Michel et al. / Geothermics 31 (2002) 169–194


8. Conclusions Hydrogeochemical and isotope data for waters sampled in the White Lake basin were interpreted in the context of basin hydrogeology to define the flow systems and to provide insight into the geothermal character of the basin. The conceptual hydrogeological model consists of three flow subsystems: shallow, intermediate and regional. Shallow, fresh groundwaters, which migrate within the surficial unconsolidated sediments, are chemically dominated by Ca, Mg, and HCO3. Isotopically they often indicate that the waters have undergone some evaporation prior to recharge. Deeper, but still local, flow systems within the bedrock have evolved chemically from the shallow groundwater either through the addition of Na and Cl ions or by Na–Ca cation exchange. Groundwaters encountered in the deep diamond drill holes are representative of intermediate flow systems, and show no chemical evidence of equilibrating at temperatures higher than about 80 C. These waters also display a uniform isotopic signature that suggests recharge under colder climatic conditions or at higher elevations than the shallower groundwater systems. Chemically, the intermediate depth waters have evolved to reflect the lithologies encountered in the subsurface. From a thermal perspective it is likely that the intermediate groundwaters have a strong influence on the thermal regime, and that perhaps they serve to reduce the gradients in the central part of the basin. Regional flow systems, which would exist at greater depths within the basin, have likely not been encountered during this study, unless significant upward vertical hydraulic gradients are present in the valleys. Certainly, faults situated in the central portion of the basin could act as conduits for flow. Geothermometry results suggest equilibration temperatures of less than 80 C for waters in deep boreholes. The current study concludes that the large number of major faults that dissect the basin and offset the stratigraphic units exert a significant effect on the hydrogeologic regimes, especially at the regional scale, and may be important considerations for the development of a geothermal resource in the basin.

Acknowledgements The authors would like to thank Drs. A. Jessop, A. Judge and T. Lewis of the former Energy, Mines and Resources Canada (EMR) for their many discussions concerning geothermal systems and for EMR’s financial support for this project. Additional funding was provided to Dr. Michel through a Natural Sciences and Engineering Research Council of Canada (NSERC) operating grant. The authors would also like to thank the Atmospheric Environment Service (AES) staff at the Penticton airport for collecting precipitation samples, and Mr. R. Drimmie of the University of Waterloo for his field assistance. Finally, they would like to express their sincere appreciation to all of the well owners who gave their permission to collect samples and for their enlightening and informative conversations about the area.


F.A. Michel et al. / Geothermics 31 (2002) 169–194

References Church, B.N., 1973. Geology of the White Lake basin. BC Department of Mines and Petroleum Resources Victoria, BC, Bulletin 61. Church, B.N. 1979. Geology of the Penticton Tertiary outlier. BC Department of Energy, Mines and Petroleum Resources, Victoria, BC, Preliminary Map 35. Clark, I., Fritz, P., 1997. Environmental Isotopes in Hydrogeology. Lewis Publishers, New York. (328 p.). Drimmie, R.J., Heemskerk, A.R., Mark, W.A., Weber, R.M., 1991. Deuterium by zinc reduction. Technical Procedure 4.0, Rev.02, Environmental Isotope Laboratory, Department of Earth Sciences, University of Waterloo (6 p). Drimmie, R.J., Heemskerk, A.R., 1993. Water by CO2 equilibration. Technical Procedure 13.0, Rev.02, Environmental Isotope Laboratory, Department of Earth Sciences, University of Waterloo (11 p). Fournier, R.O., 1973. Silica in thermal waters: laboratory and field investigations. Proceedings International Symposium on Hydrogeochemistry and Biogeochemistry, Tokyo, pp. 132–139. Fournier, R.O., 1979. A revised equation for the Na/K geothermometer. Geothermal Resources Council Transactions 5, 1–16. Fournier, R.O., Potter II, R.W., 1979. Magnesium correction to the Na-K-Ca chemical geothermometer. Geochimica et Cosmochimica Acta 43, 1543–1550. Fritz, P., Drimmie, R.J., Frape, S.K., O’Shea, K. 1987. The isotopic composition of precipitation and groundwater in Canada. In: Isotope Techniques in Water Resources Development, Proceedings Series, International Atomic Energy Agency, Vienna. pp. 539–550. Giggenbach, W.F., 1988. Geothermal solute equilibria. Derivation of Na–K–Mg–Ca geoindicators. Geochimica et Cosmochimica Acta 52, 2749–2765. Gonfiantini, R., 1986. Environmental isotopes in lake studies. In: Fritz, P., Fontes, J.-Ch. (Eds.), Handbook of Environmental Isotope Geochemistry, Vol. 2, The Terrestrial Environment, B.. Elsevier, Amsterdam, The Netherlands, pp. 113–168. Grant M.B., Michel, F.A., 1983. Study of the hydrogeology of the White Lake basin, British Columbia. Contract report for Department of Energy, Mines and Resources Canada (55 p). Jessop, A.M., Judge, A.S., 1971. Five measurements of heat flow in southern Canada. Canadian Journal of Earth Sciences 8, 711–716. Lawson, D.W., 1968. Groundwater flow systems in the crystalline rocks of the Okanagan Highland, British Columbia. Canadian Journal of Earth Sciences 5, 813–824. Lewis, T., 1984. Geothermal energy from Penticton outlier, British Columbia: an initial assessment. Canadian Journal of Earth Sciences 21, 181–188. Michel, F.A., Fritz, P., 1982. An initial study of the hydrology of the White Lake Basin, BC Contract report for Department of Energy, Mines and Resources Canada (30 p). Souther, J.G., 1977. Volcanism and tectonic environments in the Canadian Cordillera — a second look. In: Volcanic Regimes in Canada. Baragar, W.R.A., Coleman, L.D., Hall, J.M. (Eds.), Geological Association of Canada, Special Paper, Vol 16, pp. 3–24.