NOTICE CONCERNING COPYRIGHT RESTRICTIONS
This document may contain copyrighted materials. These materials have been made available for use in research, teaching, and private study, but may not be used for any commercial purpose. Users may not otherwise copy, reproduce, retransmit, distribute, publish, commercially exploit or otherwise transfer any material.
The copyright law of the United States (Title 17, United States Code) governs the making of photocopies or other reproductions of copyrighted material.
Under certain conditions specified in the law, libraries and archives are authorized to furnish a photocopy or other reproduction. One of these specific conditions is that the photocopy or reproduction is not to be "used for any purpose other than private study, scholarship, or research." If a user makes a request for, or later uses, a photocopy or reproduction for purposes in excess of "fair use," that user may be liable for copyright infringement.
This institution reserves the right to refuse to accept a copying order if, in its judgment, fulfillment of the order would involve violation of copyright law.
CHAPTER IX
Chemical Investigations at Waiotapu By S. H. WILSON, INSTITUTE OF NUCLEAR SCIENCES, LoWER HuTT.
INTRODUCTION From the chemical work on the discharges of the holes at Wairakei, it had become clear that these were fed from a fairly uniform body of chloride water at a temperature of about 250°c. After the opening of any hole there is a very variable initial period of about one year during which the chloride content is approximately 25 % lower, but finally the chloride content becomes almost constant. There are small variations from hole to hole, but it appears that these can be explained on the hypothesis that originally there was water of uniform enthalpy and chloride content, and there have been variable changes in enthalpy and chloride content by dilution with cold water, loss of heat as steam or by conduction, or drawing of excess steam into the hole discharge. On account of this variation it is best to express the other constituents in the form of ratios to chloride or to sodium.
At Wairakei it has been found that certain ratios are fairly constant over the whole field, and this .fact reinforces the deduction from the chloride content, of a body of water of uniform composition, at least originally. The ratios that are practically constant are those of chloride to fluoride, to boric acid, and to arsenic. lt. is also easy to determine these ratios accurately as all four constituents are present only in the water fraction of the discharge. There are three other ratios that' are more difficult to determine because three of the constituents are present in the discharge in both the water and the steam fractions. It is necessary to sample steam and water separately, and to calculate the composition of the original steam-free water phase. With separate sampling of steam and water, it has been difficult to eliminate various sources of error in the calculated composition of the, original water phase. The ratios of chloride to total carbon dioxide (i.e., carbon dioxide present in the steam as gas, in the water as carbon dioxide, and as bicarbonate) and of chloride to total sulphur (i.e., sulphur present in the steam as hydrogen sulphide and in the water as sulphate) seem also to be constant for the deeper bores of normal enthalpy, but there are considerable differences associated with differences in the enthalpies of the hole discharges. Where the underground water has cooled by loss of steam, much carbcn dioxide and hydrogen sulphide is carried off with the steam. Ammonia is another constituent that is determined in both water and condensed-steam samples. The ratio of chloride to ammonia is quite variable. It is low in the shallow holes just penetrating the impermeable Huka beds, but high in the holes deeper in the Waiora breecia or drawing from fissures in the ignimbrite. More useful, and more easily determined (for the constituents are both entirely
87 in the water phase) is the ratio of sodium to potassium. This is high in the waters from shallow holes and still fairly high in waters from holes tapping hot water in the "permeable volcanics" where apparently time of contact or the greater surface available has permitted potassium to replace sodium to a considerable extent in feldspar minerals. In the deeper holes in the western part of the area, where these have tapped fissures in the ignimbrite, sodium/potassium ratios have been as low as 9·5.
The realisation that the holes were fed from a fairly uniform body of chloride water led to a better understanding of the natural activity. There are two main hot-spring areas at Wairakei - the Waiora and Geyser Valleys. Waiora Valley is an area of acid sulphate and mixed chloride-sulphate waters. The waters seem to be surface water heated by steam rising up from chloride water below, although in springs in the lower part of the valley there is some admixture with chloride water. Although the area is close to the locality where the most successful holes have been drilled, the composition of the spring waters is of no help in indicating the character of the hot chloride water below. At Geyser Valley, however, the outflow of hot chloride water is very much greater. Analyses of the spring waters have shown that the ratios of chloride to fluoride, boric acid, and arsenic, are much the same as those for waters from the holes. The chloride content of the waters is somewhat lower than that in water separ- ated at amospheric pressure from the hole discharges. It could be explained by admixture with about 20% of cold water. The sodium/potassium ratio is somewhat higher than the ratio generally found in waters from holes in the permeable volcanics. lt appears that the hot chloride springs at Geyser Valley are supplied from the hot water discovered by drilling. The main changes are that a large amount of steam, and with it most of the gas, escapes from the water, some of it in the hot pools and geysers, and there is some admixture with cold water before the hot water reaches the springs. By far the largest spring is the Champagne Cauldron, which contributes about 50 % of the heat supplied to the stream draining the area. The water of this spring is less diluted with cold water than any other spring. Ifthere had been no drilling in the Wairakei locality it would have been possible to obtain a good idea of the character of the under- ground "primary chloride water" from the analysis of the water of Champagne Cauldron. Owing to the loss in the steam of carbon dioxide, hydrogen sulphide, and ammonia, it is not possible to get any information as to the content of these constituents in the underground chloride water.
When it was appreciated that in most thermal areas the surface activity is due to steam and water coming up from "primary chloride water", probably of uniform composition in each area, it became of interest to take water samples in other thermal areas in the Rotorua-Taupo region in order to check this hypothesis, and to determine what differences there were in the characteristic ratios in the various thermal areas. Sampling for this purpose was carried out in 1955. The aim was to take samples from the springs of highest flow and chloride content. Previous sampling was not always a guide, as the most striking pools or the most vigorously boiling were not always those required. In some cases, as at Geyser Valley,
88 one spring had a flow so much greater than the others that it alone was worth sampling. 1n other cases, a number of springs had to be sampled, and by judgment from the general agreement of the ratios, a decision had to be reached as to whether they were fed from one uniform body of underground chloride water. It should be pointed out that difficulties are caused by the occurrence of waters of mixed origin, i.e., those formed by mixing of chloride water with surface water heated by steam. This steam carries no chloride, a small amount of ammonia, and very small amounts of boric acid and hydrogen fluoride. If the mixed water con- tains some surface water heated by steam, the chloride/ammonium ratio can be considerably lowered, but the chloride/fluoride and chloride/boric acid ratios would be hardly affected. The fact that waters considered to be of mixed. origin have low chloride/fluoride and chloride/boric-acid ratios indicates that the water has either scrubbed the ammonia, boric acid, and fluoride from a large amount of steam passing through it, or else has leached these constituents from ground in which they had previously been absorbed from steam. These waters are generally high in sulphate, formed by the oxidation near the surface of hydrogen sulphide in the steam. Such waters must be distinguished from the true chloride waters, as the ratios of the constituents give no clue to the characteristic ratios of the "primary chloride water". The result of the investigation was to confirm the belief that in most thermal areas the activity was due to the presence underground of"primary chloride water", and to show that each area had characteristic constituent ratios, differing very considerably from area to area.
NATURAL ACTIVITY AT WAIOTAPU The samples of the waters of hot springs at Waiotapu were taken in June 1955 as part of theprogramme of sampling the springs of hot chloride water in the whole thermal region. At Waiotapu, the aim was to make a representative sampling of the active area by taking three samples of springs of high temperature and good flow. The springs chosen were well separated on a line across the area. According to later work by E. Lloyd (pp. 00), there are not many springs in the area with chloride-rich water, and the three sampled are about all that are available.
At the time of sampling, the Waikite area was considered to be a distinct chloride area, and a sample was taken from the large pool there. The analysis is included in this paper, as the Waikite area lies only 2 miles along the Waikite Road from the active fumarole area on Maunga- ongaonga, and hence may be part of the general activity at Waiotapu.
A number of analyses of waters from Waiotapu are given by L. I. Grange ( 1937), but of these only the water from Champagne Pool seems to be chloride water, and the others are all mixed or acid sulphate waters.
As the values of the constituent ratios from these analyses have not been published previously they are given in table 7A. Unfortunately at
89 the time the analyses were made, fluoride and arsenic were not generally determined, so that two important ratios are missing.
1n 1937, the author assisted Dr A. L. Day (formerly Director of the Geophysical Laboratory, Carnegie Institution, Washington) in a survey of the hot-spring areas of the Rotorua-Taupo region. At that time it was thought that the samples of most significance could be. obtained from the acid springs in the hottest part of the Waiotapu area, i.e., where steam was most in evidence. Hence all the three samples of water taken for analysis were from the flat below Bridal Veil Falls - the first flat to the north of Lake Ngakoro. It is now known that the analyses of such acid springs give little useful information, but as these analyses have not been published they are included in table 6A.
Comparisons between the various analyses are best made by means of ratios of pairs of constituents, and these ratios are given in tables 7A, B, and c. Comparison of the ratios is facilitated by a method devised by the author, in which the relations between the pairs of constituents are given in the form of a fraction which has been named the "Chemical Index". The numerator of the fraction is the content of the first con- stituent in millimoles per litre, and the denominator is the mole ratio of the first constituent to the second. It is the latter that have to be compared, but the numerators are useful, as they indicate whether there are great differences in the absolute concentrations. lf there are, good agreement in the ratios is not to be expected. A further point is that the fraction itself, calculated to a decimal fraction, gives the content consistent in millimoles per litre. For example, the chloride/ammonium chemical index for Champagne Pool in (table 7B) iS 53/199. The chloride content is 53 e.p.m., the chloride/ammonium ratio is 199 and the ammonium content is 53/199, i.e., 0·26 e.p.m.
If more than three different constituents have been determined a large number of ratios can be given, but it is considered sufficient to give the ratios of chloride to the various other anions, and the ratios of sodium to the other cations. In addition, the sulphate/ammonium ratio is useful. There does not at first sight seem to be any connection between these two constituents, but it appears likely that ammonia is formed from atmos- pheric nitrogen in solution in the ground water by combination with hydrogen in the magmatic steam. The sulphate may be formed by the oxidation of sulphur compounds in the magmatic steam by atmospheric oxygen in solution iii ground water. The calculated value of the ratio is 8·2 if hydrogen sulphide is oxidised, but it is possible that a small content of sulphur dioxide originally in the magmatic steam has been oxidised, in which case the ratio would be 2·05. The calculated ratio cannot be expected, because firstly there is evidence that about 90070 of the nitrogen goes off with steam lost from the body of hot chloride water during the long period underground, and secondly, on coming to the surface the hot chloride water loses steam till its temperature falls to 99 °c, and this steam carries with it most of the ammonia.
90 TABLE 6A: Waters from Springs in Waiotapu Hot-spring Area - Complete Analyses
Waikite, 2 Miles 1 Mile South West of Fumarole Locality On the Flats in First Explosion Crater North J 8 Mile North Centre of Main of Lake Ngakoro i of Champagne Active Area of Champagne Area of Maungaongaonga Pool F Pool
Yellow Pool Greenish-black • Green Pool I Postmistress | Champagne Boiling Pool Largest Pool Description or Name by Lake Ngakoro • 1 Pool 1 pool I Pool
20 64 97 Spring No. on Survey Near 82 Near 76 Near 75 Sheet N85/6 1 8/3/37 • 27/6/55 • 27/6/55 27/6/55 27/6/55 Date of Sampling 513137 8/3/37 to e.p.in. P.P.m. e.p.m. P.P.m. e.p.m. p.p.m. e.p.m e.p.Ill. P.P.m. e.p.m. P.p m. e.p.m. P.P.m. PAni. 1·87 478 20·78 1146 49·83 1113 48,39 180 7·83 - Sodium Na+ .. .. 404-5 17·59 422 18·37 0·21 22 0·57 160 4·09 35 0,90 8,3 Potassium K+ .. .. 73,5 1·88 11 0·28 93 2·38 0·30 4 ·3 0·62 8 1·15 5·9 0·85 2·1 Lithium Li+ .. .. 29 1·45 33 1·65 5 0·25 40 2·00 27 1·35 12-5 0"62 5 ·6 0·28 Calcium Ca '+ .. 0·18 2-4 0·20 1,9 0·16 2'5 0·21 Magnesium Mg'+ .. 7·5 0·62 3·5 0·29 1 0·08 2 ·2 Aluminium A 1 3+ . 17·5 1·96 8'9 0.99 2·5 0'27 Ferric iron Fe'+ - 5·0 0·26 8·2* 0·29* 0·0 0·00 16 6·13 0·672 21 1,16 1 ·1 0.67 '4·8 0'27 '•2·8 0·•' Ammonium NH 2 - 65·5 3·64 62 0·34 Hydrogen ion H+ 1·6 1·6 1·6 1·6 0·1 0·1 Hydrogen ion H+ (by differ- ence ·· 2·4 .. 2·3 .. 51·26 .- 8-57 29·55 7·01 ,. 23·02 22·50 56·99 Sum of cations . . 52·99 1774 50·03 114 3·22 612 17'27 ji·5 0·90 858 24·19 695 19,60 1879 Chloride Cl- . . 1·9 0·023 35 0.044 3·4 0·044 0·03 0·0004 Bromide Br- .. 0·4 0·003 2-4 0·019 1·5 0·012 0·033 0·0026 0·87 0·046 Iodide 1- - 6·8 0·36 4·05 0·21 5·4 0·28 Fluoride F- .. .. 1·48 99 2·07 5.9 1·23 32,4 0·76 Sulphate SO :2- 594 12 36 306 6·j7 99·5 2,07 71 Bisulphate HSO 4- . , 72 0 75 41 0·42 TABLE 6A: Waters from Springs in Waiotapu Hot-spring Area - Complete Analyses-continued
Locality Waikite, 2 Miles ., On the Flats in First Explosion Crater North f Mile North Centre of Main of Lake Ngakoro * Mile South West ofFumarole of Champagne Active Area of Champagne Area of Pool Pool Maungaongaonga Description or Name . • Yellow.Pool • Greenish-black Green Pool Postmistress Champagne Boiling Pool Largest Pool pool Pool Pool by Lake Ngakoro
Spring No. on Survey • Near'82 Near 76 Near 75 • Sheet N85/6 20 64 97
Date of Sampling 5/3/37 8/3/37 8/3/37 | 27/6/55 27/6/55 27/6/55 27/6/55 •.0 1 w Bicarbonate HCO 3- .- 1 .. .. 69 1·14 233 3 81 394 6·46 318 5·22 Carbonate CO 32- .. 1 1·8 0·06 Bisulphide HS- .8 4·3 0·13 0·10 Sum of anions .. • .. 30 •38 .. 7269 1 0·61 6·05 0·661 26·26 22-37 59·01 5779 .. 9·20 Metaboric acid HBO 2 ,· i 41 0·94 10 0·23 34 0·78 29·6 0·66 11 Metasilicic acid. H :Sio , 1 482 6·17 2·52 101 2·29 5,7 0·13 363 4·65 587 7'52 244 3-12 500 6'40 Carbon dioxide CO 2 ·· 1 ·· 328 4·20 160 2·05 .. 0 96 209 113 2·57 Hydrogen sulphide H2S 1 0·0 - 9·5 0·2 24 0 55 8 .. 0,1 16 0·47 2·8 0·08 0·1 Arsen•us acid HA,O, . - .. 2.8 - ' 2'8 ' 1·65 0·815 8 0·074 11·4 0·106 0·38 0·0035 3·7 8·77 6·50 6·65 7,22
*Iron in ferrous state Fe 2+ p,p.m. = parts per million. e.p.m. = equivalents per million = milli-equivalents per litre. TABLE 6B: Waters from Springs in Waiotapu Hot-spring Area - Partial Analyses
., •1 Mile North-west 1 * Mile North-west • i Mile North-east of 64 Locality of 64 of 64 1
. . Postmistress Pool Pool in Crater Champagne Pool ' Lady Knox Geyser • Hot Spring Near Description or Name NO. 9 • Lady Knox Geyser
52 Spring No. on Survey Sheet N85/6 20 61 64
24/7/58 to 23/1/59 Date of Sampling 23/10/58 23/1/59 24/7/58 to and 23/1/59* 23/1/59t 23/1/59t •0 SJ e.p.ni. P.P.m, e.p.m. p.p.ni. e.p.m. P.P.m. e.p.m. p.p.m. e.p.m. P.P.tri. 49·4 500 21·7 1 200 8·7 Sodium Na+ .. .. . 1 445 19·30 1090 47·4 1137 155 3·96 160 4·09 78 2·00 32·5 0·83 Potassium K+ . 1 19·5 5·0 0-17 3·8 054 8·2 1·18 1 8·2 1·18 5.1 0·73 1·2 Lithium Li+ .. 0·53 | .. .. Calcium Ca 2+ . | 8·2 0·41 30'7 2·8 0·23 1 .. Magnesium M g 4 - 0·6 0·05 11'3 0·63 1 .. Ammonium N H 4+ . - 0·4 0·02 . 1 684 19·3 1766 49·8 1961 820 23·1 270 7·6 Chloride Cl- 5·2 4 27 Fluoride F- .. .. 1 6·1 0·32 i 5·2 O·27 1 2·86 1 Sulphate SO i 2- . 1 93 1·94 143 110 2·51 • 11·3 0,26 Metaboric acid HBO 2 .- 28·2 0,64 12·0 0·27
*Mean of two analyses. t Mean of four analyses. i Mean of three analyses. TABLE 7A: Chemical Indices* from Analyses of Springs at Waiotapu Given by Grange ( 1937)t in Bull. 37
*See note at end of Table 7c. tDescription of springs 31-43 previously described by Grange (1937) with numbers on sheets N85/6 or 85/3, are as follows:
Near Champagne North of Near Lady Knox Maungaonga- 1 Areas 1 *to 8 Miles Pool Champagne Geyser onga Thermal South of Champagne Pool A rea Pool Number in Bull. 37
1 31 • 32 • 33 • 34 35 I 36 • 37 38 39 40 • 41 42 43
Chemical Index ( millimoles per kg/mole ratio) 1.1 Chloride 56 24 0361235 20 4 14 .. 1·2 5 2I 1·1 2 lehloride-ammonium ratio /26 /19 / 1205 1 /48 290 1 1- /- 1 /20 113 /33 /chloride-sulphate ratio , , /5 /24 144 1 145 I /29 119 /2 11'8 /0' 1 /130 IG t14 /chloride-boric acid ratio 1 /23 /52 116 156 f54 17 /20 Sulphate • 03610·04 1·2 0,8 38 1 1·1 1 2 7·8 2 3·5 0·05 'sulphate-ammonium ratio /0·6 /0·6 /1·1 / /9 1 15 1- \ 1- /10 /- /0·15 /4 /1-4 Total anions 59 26 25 1 7 22 9 5 6 37 6 /total anions-bicarbonate ratio /85 1 /79 /160 | /75 , /110 1 60·8 /2 2 /5-5 /4-3 | /1-2 Sodium 53 22 17 24 1 5 2() 7 14 1 0·5 20 27 | 8-5 /6·6 • /sodium-potassium ratio /12·6 /11 /26 /9 /23 /35 1 19 13*5 /1·3 /17 /12 | /23 /sodium-calcium ratio . . /55 i /62 | 137 '16 /18 /74 /ll /16 /4 1 /7 /14 /98 ' /29 0·5 • 0·3 0·9 0·2 0·1 ' 1·4 0·3 0·3 0·3 Calcium 1·0 0·4 0·3 0·6 /5 /27 •14 | /0·8 /56 /56 /0·9 /calcium-magnesium ratio 180 130 /2 /13 n fl 6·6 3,5 6·6 PH .. | 4·9 5·6 ,. 1 7·2 5·7 6·2 34 1 ..
31. Champagne Pool. N85/6/64 33. Yellowish spring near creek draining from Champagne Pool. Probably N85/6/72. 32. Acid spring below Primrose Terrace. 35. Slightly acid spring, 10 chains north of Venus Bath. N85/6/46 or 47. 34. Spring beside Waiotapu River, 10 chains upstream from bridge. 37. Boiling neutral spring at ford on Waiotapu River, 1 mile south-east Probably N85/6/55. of Waiotapu Hotel. N85/6/54 or 119. 36, Weakly acid spring, 35 chains north-east of Venus Bath. N85/6/32 39. Acid pool, 57 chains along track, now a road to Waikite Valley or 122. Probably N85/6/6. 38. Acid spring close to previous spring. N85/6/54 or 119. 41. Weakly acid spring in swamp on Mr Hickey's farm, 20 chains wesi 40. Turbid acid spring, south-south-east of flank of Maungaongaonga. of Rotorua-Taupo road at 22 mile post, 1• miles south-west of Probably N85/3/74 or 75. Champagne Pool. 42. Acid spring between Maori Bath and mud pot at Otakitaki, 41 miles 43. Weakly acid spring, southernmost of group, 8 miles south-west of 0 south-south-west of Champagne Pool. Champagne Pool. L.,1 TABLE 7B: Chemical Indices* from Analyses Given in Table GA
1 * Mile On Flats in First Explosion Crater 1 North-west I Locality Of • 3 Mile North of Lake Ngakoro South Champagne ofid. Pool
Champagne Yellow Green- Green Post- Champagne Boiling Waikite Cauldron Description or Name Pool black pool mistress Pool I Pool by • Large Wairakei Pool Pool Lake Pool (for Ngakoro Comparison)
Spring No. on Survey Sheet N85/6 • Near 82 | Near 76 • Near 75 • 20 64 97 t<) 0 Chemical Index (millimoles per kg mole ratio) Chloride 17 0·9 24 20 53 50 3·2 50 /chloride-ammonium ratio /4·8 11-6 /21 /288 /199 /320 124 /1200 /chloride-fluoride ratio /54 /252 /176 /70 /chloride-bromide ratio /164 /840 /1200 /1260 /800 /1000 Bromide .. .. 0·02 0·044 0·044 0·003 0·05 /bromide-iodide ratio 18 8.4 13·6 /0·14 9 Chloride - 17 0·9 24 20 53 50 3·2 50 /chloride-sulphate ratio 12·2 40·3 f23 126 151 /82 19·6 /188 /chloride-boric acid ratio /18 i4 /31 129 1•21 722 i24 /chloride-arsenic ratio /24·5 /1260 7720 1 /480 /920 /1020 *h Sodium 18 2 18 21 50 48 7·8 46 /sodium-potassium ratio /9-4 16,7 11·1 86 •12·2 154 fyI /17·8 /sodium-lithium ratio /33 143 M7 126 429 ,/sodium-calcium ratio .. /i 8 ,/2 - 8 157 /148 169 159 i61 171 Calcium ...... 1 0·7 0·3 0·1 0.7 0·8 0·12 0·7 /calcium-magnesium ratio . . /3·2 /4·6 /7·8 1\·5 41.3 /10·5 /1 ·2 /13
Total anions...... 23 59 58 9 52 /total anions-bicarbonate ratio. . 110 /16 19 12 /42
Sulphate 6·2 3·2 1·0 1·5 1·0 0·6 0·4 0·26 /sulphate-ammonium ratio . . 11 7 /10 10·9 122 13·9 ,/3·9 25·3 /65 \C> \] *See note at end of T'able 2c. I'ABI,E 7C: Chemical Indices* from Analyses (;iven in '1':,ble 68
Locality . , • : Mile North-west | 1 Mile North-west I 1 Mile North-cast • Of 64 of 64 1 of 64
Description Or Name - .. Postmistress Poolin Champagne Lady Knox Geyser Hot Spring Near Pool Crater No. 9 Pool Lady Knox Geyser
Spring No. on Survey Sheet N85/6 20 61 64 52
Date of Sampling . . . | 23/10/58 and 23/1/59 24/7/58 to 24/7/58 to 23/1/59 23/1/59+ 23/1/59* 23/1/59• Chemical Index ( millimoles per kgt/ mole ratio)
Chloride 19 50 55 23 8 • /chloride-ammonium ratio /880 /50 /88 ,/59 /chloride-fluoride ratio /60 /202 /84 /chloride-sulphate ratio 120 /31 /chloride-boric acid ratio /30 /i 82 111 /24 119 Sodium 19 47 49 20 9
/sodium-potassium ratio 139 1'/2 /12 /11 /10·5 /sodium-lithium ratio 86 140 /42 /29 /50 /sodium-calcium . . /94 164 Calcium 0·2 0·8
/calcium-magnesium ratio /8. 16·5
*In order to facilitate comparison of different waters the relations between various pairs of constituents are expressed in the form of fractions, of which the numerator is the content of the first constituent in moles per million (millimoles per litre) and the denominator is the mole ratio of the first con- stituent to the second. It is the denominators, the ratios, that are to be compared, and similar waters should have the same ratios independent of small variations in absolute concentrations, but attention should also be paid to the numerators, for if the absolute concentrations differ greatly good agreement in the ratios is not to be expected. The fraction is called the "Chemical index". tMean of two analyses. *Mean of four analyses. §Mean of three analyses. DISCUSSION OF THE ANALYSES OF HOT-SPRING WATERS
The analyses of the waters are best compared by means of the con- stituent ratios given in tables 7A, B, and c. The ratios calculated from the analyses given by Grange (1937) are not of much use as there are no figures for fluoride and arsenic, while boric acid was not always determined. The data for acid-sulphate waters low in chloride are of no use at all. Figures for mixed chloride-sulphate waters must be used with care. For instance, the low figures for the sodium/potassium ratio for samples 33 and 35 in table 7A (chemical indices 17/6·6 and 5/9) are evidently due to leaching by acid water near the surface, whereas the ratio 12·6 for Champagne Pool may be near the true value for the underground chloride water. The low values for the chloride/ammonium ratio indicate heating of water by steam, not magmatic steam, but steam separating from "primary chloride water" at depth. The high figures for the chloride/boric-acid ratio for some springs with consider- able chloride seem rather doubtful. Lloyd (pp. 00) does not report any springs with high chloride/boric-acid ratios. Such ratios would indicate that boric acid was being deposited or absorbed in the ground. Instead, boric acid is being taken up by water from ground into which it has been carried by steam.
The most interesting point from table 7A is that water with fairly high chloride content, and rather similar to the water of Postmistress Pool, is found as far away as 4 miles south of Champagne Pool (Grange, 1937). The water seems to have picked up some ammonia from steam or from the ground.
From the analyses and "chemical indices" for the samples collected in 1937 (tables 6A and. 7B), it .is evident that although they were taken as representative of the springs of the hottest part of the area, the figures are not of great significance. The second sample is of water heated only by steam near the surface, and is of no use. The first sample is fairly high in chloride content, but the chloride/ammonium ratio shows that it has been heated also by steam. The third sample is the only one that seems to be of much use. On the basis of the chloride/boric-acid, and chloride/ sulphate chemical indices, 24/31 and 24/23, it seems to resemble Post- mistress Pool. It is unfortunate that because of lack of data the chloride/ fluoride ratio is not available, but it seems unlikely that the water can be from the same source as Postmistress Pool, as the spring is just to the south of Champagne Pool. Once again, it is hard to see how the chloride/boric-acid ratio can have been raised.
lt is apparent from tables 6 and 78 that the water of Champagne Pool is very similar to that of Champagne Cauldron at Wairakei. As in both cases the springs mentioned have by far the greatest flows in their respective areas, the composition of the water of Champagne Pool may be taken as very similar to that of the "primary chloride water" underlying and supporting the activity at Waiotapu. The most marked difference from Wairakei water is in the sulphate content, which is considerably higher at Waiotapu. lf the sulphate has been formed by the oxidation of hydrogen
99 40 sulphide, there could not have been enough oxygen dissolved in ground water for this oxidation. Hydrogen sulphide in steam or gas coming to the surface can be oxidised by oxygen in the air to sulphuric acid, and this can be carried down deep by rainwater percolating through the porous ground. If there is at Waiotapu, as there is at Wairakei, a large body of chloride water of uniform composition, it is hardly likely that the ground water, which formed this chloride water by mixing with a small amount of magmatic water, entered the earth just where hydrogen sulphide was being oxidised at the surface. lt is known, however, that at Wairakei, the upper layers of the chloride water are diluted by ground water of local meteoric origin, and these diluted layers disappear as the bore discharges continue. If this local ditution also occurs at Waiotapu, the high sulphate content may have a local origin. in that case, it would be expected that the sulphate content will decrease with time to a value as low as that at Wairakei. If this does not occur the other explanation is that the sulphate comes from the oxidation of a small amount of sulphur dioxide. This compound is in equilibrium with hydrogen sulphide at high temperatures, and a small amount may still be in the magmatic steam when it mixes with ground water.
The sodium/potassium ratio at Wairakei is not constant, and varies from 30 to 9·5. High values are obtained with bores tapping water from "permeable volcanics", whereas the low values are characteristic of water from the high output bores fed from fissures in the ignimbrite. The low value for Champagne Pool indicates that the water is of deeper origin than that at Geyser Valley, and that it may come more directly from a fissure in more resistant rock.
The chloride/bicarbonate ratio is also lower for the water of Champagne Pool. This may indicate either that the water has been in long contact with the country rock at a higher temperature than at Wairakei, or that the content of carbon dioxide was greater. The evidence from the sampling of the discharge of Hole 3 is that the gas content is somewhat higher than at Wairakei. There is an apparent contradiction between the evidence from the bicarbonate, and from the sodium/potassium ratio, but it is probable that the replacement of sodium by potassium in minerals in the rock is favoured by lower temperatures.
The chloride/arsenic ratio is lower in the water of Champagne Pool, and this lower ratio is probably a characteristic of the underground water at Waiotapu. Grimmett and Mcintosh (1939) discovered that the poisoning of cattle at Reporoa, to the south of Waiotapu, was due to arsenic in the water draining from the country there. In view of this it would have been expected that the arsenic content of underground water at Waiotapu would have been considerably higher than at Wairakei.
Comparison of the chemical indices for Champagne Pool, and the spring by Lake Ngakoro, indicates that they are probably fed from the same body of water. As the outflow from the latter spring is small, it cannot be expected to represent the composition of the underground chloride
100 water as well as does the water of the much larger spring. The high sodiumlpotassium ratio indicates that the water of the lake spring has travelled a longer path, or has been standing stagnant in porous beds, perhaps at a lower temperature. The arsenic content is appreciably higher, and this may be significant in relation to the high arsenic content at Reporoa. It is possible that the water is picking up arsenic previously deposited in the ground.
Comparison of the results for Postmistress Pool and Champagne Pool indicates that the waters are quite difTerent. The chloride content is less than half that of the latter, and the various ratios are quite different. Moreover the differences do not seem to be consistent with the possibility that the water of the Postmistress Pool originates partly by mixing of "primary chloride water" with surface (meteoric) water, and partly by heating by "secondary steam". It is true that the low chloride/fluoride ratio is consistent with such a mixed origin, but the higher chloride/boric- acid and chloride/ammonium ratios are quite inconsistent with this. The respective chemical indices are 20/29 and 20/288, as compared with 53/21 and 53/199. In addition such mixed waters are usually high in sulphate content, and in this case the chloride/sulphate ratio is somewhat higher, but the absolute content of sulphate is actually lower (chemical indices 20/26 and 53/51 ). At the time of analysis, the composition seemed to be anomalous, but no great significance was attached to this one spring with small output. The composition might have been due to some accidental, very local, circumstance. It was only when analyses of waters from bores became available that an explanation of the composition became possible.
The water of the spring at Waikite is hardly a chloride water at all, for the bicarbonate content is higher than the chloride content. This chloride content is, however, still too high for the assumption that the chloride is leached from the ground, or brought in by surface water, and it must be taken to be of magmatic origin. The low chloride/fluoride and chloride/ ammonium ratios point to heating by steam, although the chloride/boric- acid ratio does not support this (chemical indices 3·2/70, 3 ·2/24, and 3·2/24). The sulphate content is not high, so that the water is not a mixed chloride-sulphate water. The composition is approaching that of the bicarbonate water found by the prospecting holes at the west end of D-line, Wairakei (Wilson, 1955, p. 37). The explanation in that case was that ground water was heated by steam coming off from "primary chloride water" deeper down. This steam had, however, lost its hydrogen sulphide content, probably by formation of pyrite (iron sulphide), so that by attack of the hot water on the underlying strata bicarbonate was formed. At Waikite it appears that the water heated by steam had already taken up some chloride from a magmatic source. As the spring at Waikite is the main one in the area with a fair flow, the composition of the water must be given considerable significance, as representing a large body of under- ground hot water. However, it cannot be decided from the analysis, even with the aid of the ratios, whether the area has any relation to the activity at Waiotapu.
101 PRECIPITATION OF CALCITE An account of the factors influencing the formation of deposits ( silica and calcite) in drillholes has been given by Ellis ( 1961). The following discussion of the precipitation of calcite. particularly in the holes at Waiotapu, is taken from his work. The solubility of calcite can be expressed by the following equation: (Ca2+) ( HCO-3)2 - YPC02· At 250'c in 0·05 molar sodium chloride solution with concentrations expressed in p.p.m. and pressure of carbon dioxide in atmospheres. the value of Y is 1 ·7, and at 200 °c, 14. 1 fthe concentrations of calcium and bicarbonate are known by analysis one can calculate the minimum pressure of carbon dioxide, PcO 2 (min). necessary to prevent calcite precipitation. In the water underground all the carbon dioxide is in solution and, as the discharge passes up the hole, separation into water and steam occurs with most of the carbon dioxide passing into the steam phase. In any calculations. the concentra- tions of calcium and bicarbonate must be adjusted to allow for the separation of steam. The bicarbonate concentration must also be corrected for the change in the ionisation of boric and siIicic acids owing to the variation in the pH of the water with the loss of carbon dioxide. 11 is possible to calculate the partial pressure of carbon dioxide in the underground water from the carbon dioxide content of the total discharge. and the solubility of carbon dioxide. Pressures found are similar in value to PcO 2 (min) (calculated for these conditions), which shows that the waters are close to saturation with calcite. The following values of Pco 2 (min) are calculated for atmospheric separation.
COY Temperature 1 ( Moles/ 1 HCO- 3 1 Ca 2- Pco: t • 100watermoles) •I p.p.m. p.p.m (min)
Bore 6, Waiotapu 235 0·032 130 1 9 3·2 Bore 7, Waiotapu 240 0·103 160 I4 1·5 Bore 28, Wairakei 265 0.026 85 10 2·9 Bore 46, Wairakei 255 0·011 65 4 1·0
There is evidence from water samples taken under high pressure from the discharge pipes of drillholes that the carbon dioxide concentration in water flashing in the pipe may be several times that expected for equili- brium conditions. The effect of this is to make precipitation more likely. but also makes it more difficult to judge from analytical data whether such precipitation may occur. However. the tendency to precipitate calcite would be greatest for those waters with the highest values of Pco 036(min), as calculated for atmospheric separation.
102 In general Wairakei holes have been free from calcite precipitation. but some troublesome precipitation was experienced in Hole 28, Wairakei. High values of Pco 2 (min) correlate with examples of precipitation in the casings of Hole 28, and of holes at Kawerau. It would appear that trouble with deposits of calcite would also be likely at Waiotapu, and this has indeed been found to occur (see Dench, pp. 119-26, Holes 3,6.7).
GAS FROM POOLS AND FUMAROLES Samples of gas are easily collected from bubbling hot pools. So far, the composition of the gas has not been of much use in studying a particular thermal area. It would be expected that the ratios of carbon dioxide to hydrogen and methane would be of use in indicating the losses of gas from any body of hot chloride water. Owing to the differences in solubility, hydrogen and methane should tend to come off before carbon dioxide. but the temperature at which loss occurs is probably too high for solubility differences to have much effect. There is not enough experience yet to judge the significance of differences between various areas in contents of hydrogen and methane as well as of nitrogen. The latter is derived from air dissolved in ground water, or is drawn in locally from the atmosphere. Where the activity is very old, or even where the steam causing the surface activity has had to travel far, the hydrogen sulphide may be very low. and the ratios of carbon dioxide to hydrogen and to methane quite high. lt is unfortunate that there has been no opportunity to sample the gases at Waiotapu since the geothermal work began. The only analyses available are those of samples collected by the author for Dr A. L. Day in 1937. The interest was then in the content of hydrogen sulphide, and complete analyses Were not always made. The analyses are given in table 8. It will be seen that the hydrogen sulphide content is considerably higher than that found in pools in the Waiora Valley - an acid area at Wairakei (Wilson, 1955, p. 28) - and is much the same as the content at Rotokawa, a mixed area. There is also an appreciable content of other gases, and the indication is that the steam has not travelled far from its source in hot chloride water. Messrs W. J. McCabe and J. R. Hulston, Institute of Nuclear Sciences, D.S.I. R., have kindly permitted the use of a recent analysis made by them (table 8 (Spring 64) ). This analysis seems rather anomalous, as nitrogen, methane, and hydrogen are unusually high. relative to carbon dioxide. However, the author collected a gas sample of very similar composition from a pool below Bridal Veil Geyser at Geyser Valley, Wairakei. The gas from Champagne Pool corresponds to a gas of normal composition. from which '70 % of the carbon dioxide has been removed. If the gas was just that associated with the "primary chloride water" which had lost steam on reaching the surface, then it would be expected that the less soluble gases - hydrogen, methane and nitrogen - would have been lost to a greater extent, and the remaining gas would contain even more carbon dioxide. Gas from an intermediate source (even though the ulti- mate source is the same) is evidently bubbling through the water, and much of the carbon dioxide has been taken into solution.
103 TABLE 8: Analyses of Gases from Hot Pools at Waiotapu Mainly Collected in March 1937
Locality First Explosion Crater North of Lake Ngakoro North of Centre of Champagne PooI Main Active Area
Hot Pool Yellow Pool Greenish-black "Frying pan" Large green pool Poe.110 chains up 5 ftby 3 ft Champagne Pool pool with pyritic 10 ft long by 2-3 ft 15 ft diameter on from footbridge collected 27/4/58 beside the large scum, 10 ft east wide in middle of eastern edge of 4 ft above pool in crater by W. McCabe of Yellow Pool flat flat Waiotapu Stream, and J. R. Hulston • pool 12 ft diameter and analysed rnainly by mass spectrometer
Spring No. on Survey Sheet N85/6 .. 81 76 75 57 64
Carbon dioxide (%)..1 86'4 88.·5 86,4 76·6 78-1 Hydrogen sulphide (%) 74 3·7 3·25 4·45 3·75 1·95 Hydrogen .. .. 1·25 1·6 7 Methane . . 1·9 1'9 Nitrogen .. ., 5 5·1 5·65. 14 Sampling of steam from fumaroles is of greater value than sampling gas from hot pools, because the content of gas in the steam can be found. This is of importance as being one method of distinguishing steam of direct magmatic origin from steam derived from "primary chloride water", i.e., "secondary steam". In the former case, the steam has a gas content of about 3 % by volume, whereas the "secondary steam" has a gas content of perhaps 0·05%, as at Karapiti (Wilson, 1955, p. 29). High gas content has been found at White Island, where the steam is expected to be magmatic, and at Ketetahi Hot Springs on Mt. Tongariro. However, the main fumarole on Maungaongaonga was sampled by L. R. L. Dunn in 1936 (Grange, 1937, p. 109). The gas in the steam was 0·13% by volume. Although the technique of sampling fumaroles was not good then, this result should be fairly reliable, as the content found by L. R. L. Dunn at Karapiti was 0·035%, compared with 0·045% found by the author in 1954. Although the gas content at Maungaongaonga is appreciably higher than at Wairakei, it is still so low that it must be concluded that the steam at Mauneaoneaonea is of secondary origin.* Although the gas content of the steam from Karapiti Blowhole was very low, that of fumaroles to the west of Geyser Valley was 0.5%, and that of the steam from nearby holes was also about the same. It would be desirable to know the gas content of the steam from fumaroles and steam vents in various parts of the Waiotapu field, and any inferences from this should also take into consideration the ratios of carbon dioxide to hydrogen and methane.
SAMPLES FROM HOLES
Collection At the time of first sampling, there was only one hole discharging - Hole 3; two other holes had been drilled but were not discharging, Holes 1 and 5. It was therefore desirable to obtain water samples from the bottom of these holes. A sampling bottle had been used previously at Wairakei, but it was doubtful whether the valves in this bottle remained closed till the bottom was reached. The author pointed out the objections to the obvious method of construction, of having a valve at the top of the bottle which was opened when the weight of the bottle was taken up by contact with the bottom of the hole. The main objection was that air in the bottle would mix with the water entering. A bottle was therefore designed by Mr P. Bangma, Ministry of Works, Wairakei, so that contact with the bottom of the hole opened a valve at the bottom of the bottle, and uncovered holes at the top. In order to ensure that the bottle would be filled only with water from the bottom of the hole, it was filled with kerosene before lowering it into the hole. At the bottom the valve opened, and the kerosene was displaced by hot water. On raising the bottle to
*A sample ofsteam and gas from the fumarole was taken by the author on 1 April 1960. The gas content of the steam was 0·29 %. The analysis of the gas was: CO 2 , 94·0 %; H :S, 3·9%; 112,0'14%; CH ,, 0·46%; N 2, 1·5 %.
105 the surface the water lost steam by flashing as the pressure fell. If there was no conductive cooling near the top of the hole the composition would be the same as that of water collected at atmospheric pressure from the discharging hole. There will always be some conductive cooling. and the chloride content will always be somewhat low: only the constituent ratios obtained from the analyses will be really accurate. A complete sampling of Hole 3 was carried out by the author in the week 2-7 September 1957. The enthalpy of the discharge was also deter- mined by a method requiring the measurement of the gas content of the steam at two pressures. Hole 1 was sampled on 7 September 1957 and Holes I and 5 were sampled by Mr A. L. Hilton on 20 September 1957. Samples from Hole 6 and from two hot springs were collected by Mr W. A. J. Mahon in 1958.
The analyses of the various samples collected are given in tables 9A. 94 and 11, and the values of the constituent ratios are given in tables I OA. 108, and 12.
DISCUSSION OF THE ANALYSES OF HOLE SAMPLES From the analysis of the water of Champagne Pool it would be expected that the holes would tap water somewhat higher in chloride and with similar constituent ratios, but not one hole has as yet yielded such water. Hole J: Hole 1 is of interest as it is very near an area of fumaroles and steaming ground on Maungaongaonga. There is a similar area at Karapiti, about 2 miles south-west of Wairakei. From the low gas content of the steam from Karapiti Blowhole, a large fumarole in this area, and from the analysis of the gas, it is concluded that the steam is "secondary steam", i.e. that it is escaping from "primary chloride water". If the primary chloride water is losing 10% as steam, and its original chloride content was 1,320 p.p.m. (Ellis and Wilson, 1955), then the flow of chloride carried by the water supplying the Karapiti area would be 1,800 g/sec. This is to be compared with a total output of the Wairakei area of 800 g/sec (Ellis and Wilson, 1955). Two holes -12 and 13- have been drilled near Karapiti, and recent work by the author has shown that the discharge of Hole 13 is Wairakei water which has lost much heat by loss of steam.
There is the further point that at Wairakei the prospecting holes at the west end of D-line tapped water low in chloride (less than 3 p.p.m.) and high in bicarbonate (Wilson, 1955, p. 37). It was evident that this was a small supply of near-surface water, heated by steam from chloride water deeper down.
The questions that can be asked in regard to Hole 1 are: (1) Is the water sampled at Hole 1 similar to the water at Waikite, and hence in- dicative of a large reservoir of such water? (2) ls this water partly heated by steam, or is it produced by dilution of hot chloride water with cold water? (3) Does the steam escaping from the fumarole area on Maunga- ongaonga separate from this low chloride water. or does it come from high chloride water at much greater depth ?
106 TABLE 9A: Analyses* of Water Samples from Holes at Waiotapu, Samples Taken in 1957
5 1 3 Hole No. 20/9/57 119157 10f9/51 419f51 Date of Sampling
e.p.rn. 1 P.P.m. e,p. m. P.P.m. e.p.m. 1 P.P.m. e.p,m. P.P.m. 33·5 20·0 1 535 23,26 770 460 20·0 460 1·5 Sodium Nat ,. ,, 27 0,69 38 0,97 1 59 1,61 3·2 0·46 Potassium Kt .. .. 0·3 F 2·3 0·33 0·6 Lithium Lit ,. - +19 1 1 0·05 Calcium Ca 21 .. 0·08 Magnesium Mgn • 0·42 0·023 Ammonium NH 4+ 24·96 Sum of cations . . 18·98 949 26 8 206 5·81 j 278 7-84 : 673- 0·21 _ Chloride Cl- .. 1 5·5 0·29 2·9 0·15 5·5 0·295 4·0 c Fluoride F- 1 83 1'72 Sulphate $0 4 2- 900 (by difference) 756(by difference) 1 37 0·61 Bicarbonate HCO r 0,05 Carbonate CO, 2- 1 1:,5 0·40 Bisulphide HS- 22-06 0·66 44' 1,00 Sum of anions . 1 Bil 1·85 1 14 0 32 29 Metaboric acid HBO 2 330 4·23 Metasilicic acid H :SiO, . 0·01 Arsenious acid HAsO 2 Analyses on Samples as Collected 0 0 0 8 Sampling pressure lb/in. 1 gauge 0·00 Free carbon dioxide 0·4 6.01 Free hydrogen sulphide 7·2 8·8 7·6 pH .. ··
*Analyses made on or calculated to samples separated at atmospheric pressure. TABLE 9B : Analyses* of Water Samples from Holes at Waiotapu, Samples Collected in 1958 and 1959
Hole No. 4 6 6 6 7 Date of Sampling 10/7/59 22/9/58 15/10/58 23/1/59 1°f7159
P.P.m. e.p.m. P.P.m. e.p.ni. p.pm. Sodium Na+ .. .. 1170 e.p.ni. p.p.m. e036p.m. P.P.m. Potassium K+ 50·9 706 30·7 587 25,5 e.p.ni. 200 5·12 860 37·4 765 33·3 Lithium Li+ .. .. 103 2-63 104 2·66 158 10·4 1·50 5·9 4-04 87 2'23 Calcium Ca 20 - 0·85 4·4 0·63 6·6 0·95 9·7 0-48 6·1 0·87 Magnesium Mg21 . 8·8 0·44 14-3 0-71 3·7 2·4 0·19 1·85 Ammonium NH 1+ · 042 3·7 0·30 3·7 3·0 0'2 O:01 0·2 0·01 S Chloride Cl- . . 1972 55·6 1063 0·9 0·05 00 Fluoride F- .. 30·0 1121 31·6 1447 40,8 Sulphate SO, 2- 5·5 0,29 9·8 0·52 2·8 1260 35·5 0,15 4-9 0,26 7·3 0·38 Bicarbonate HCO-, 89 1·85 73 1·52 32 0·52 110 52·3 1,09 Bisulphide HS- 1·80 112 1'84 116 18 0·54 j 1·90 147 2·41 Metaboric acid HBO, . 16 0·48 16 0·49 78 1·78 2 34 0'78 39 11 0·33 Metasilicic acid H :SiO 3 0·88 56·5 li29 60'7 1·38 Arsenious acid HAsO 2. 042 459 5·88 472 6·05 3·3 0·031 .5.0 0,046 Analyses on Samples as Collected Sampling pressure lb/in 2 gauge . . 0 35 Free carbon dioxide.... 25 0 0·00 .. 20 Free hydrogen sulphide . . 0·00 .. 0·00 ., 0·25 0·01 pH at 20' c .- 0·4 0·01 0·4 0·01 0.4 0·01 8·8 0·4 0·01 8·8 8·8 8·7
*Analyses made on or calculated to samples separated at atmospheric pressure. Unfortunately the two analyses of the water of Hole 1 do not agree at all. It appears from the low chloride/fluoride, chloride/boric-acid, sodium/potassium ratios, and high sodium/lithium ratio, that the first sample is more like water of shallow origin which has taken up fluoride and boric acid from the ground where it has been left by steam. It is probable that the second sample is more representative of a large body of hot water underground. In regard to the first question, it seems that the water could be the same as that at Waikite, but the evidence is insufficient. In regard to the second question the only evidence is the high bicarbonate content, which points to heating by steam. The water would then be intermediate in character between a chloride water and the bicarbonate water found at Wairakei at the west end of D-line. It seems from the low gas content of the steam that the fumaroles at Maungaongaonga are due to "secondary steam", but there is no evidence as to whether it comes from the low chloride water, or from some deeper water high in chloride. Hole 3: Hole 3 has given a discharge of low enthalpy, with gas content similar to the discharges of Wairakei holes. However, in chloride content and in chloride/fluoride and chloride/boric-acid ratios (chemical indices 19/65 and 19/29), the water of Hole 3 seems similar. not to Champagne Pool, but to Postmistress Pool. As the chloride/boric-acid ratio is higher than that for Champagne Pool, while the chloride/fluoride ratio is much lower (chemical indices 19/24 and 19/65), the contrast with water of Champagne Pool cannot be explained by admixture of "primary chloride water" with steam carry- ing fluoride and boric acid, but not chloride. The ratios seem to be charac- teristic of a different body of "primary chloride water" of difTerent origin. Hole 4:The sample of water from this hole is the only one with chloride content as high as that of Champagne Pool. It is not clear whether this result has much significance, as the bore was bleeding and not blowing at the time of sampling. Hole 6: The discharge of Hole 6 was sampled towards the end of 1958, but the analyses showed that water as high in chloride as Champagne Pool has not yet been obtained. From the ratios given in table lOA it appears that the water, although considerably richer in chloride than that from Hole 3, still resembles more the water of Postmistress Pool. The marked difference is that the sodium/potassium ratio of Hole 6 is low. This would indicate that the water comes from a deeper source than that of Hole 3. The ratios on the whole discharge given in table 12 indicate that the marked difference from the discharge of Hole 3 lies in the higher content of sulphur present as hydrogen sulphide. There is a possibility that, after a long period of discharge, layers of water of altered com- position overlying the main body of water will be exhausted, and the composition will gradually approach that of Champagne Pool. Another possibility is that Hole 6 is fed by water coming up a fault to the west of the fault feeding Champagne Pool ( as indicated by fig. 4), and that this water is of different composition from those supplying either Champagne Pool or Hole 3.
109 TABLE lOA: Chemical Indices* for Analyses Gi,en in Table 9A ( Analyses Made in 1957 and 1958)
Largest • Postmistress Champagne Bore No. Pool, Pool, 1 Waikite Pool, 3 5 Waiotapu Waiotapu (For (For Comparison ) ( For Comparison ) Comparison)
Date of Sampling 7/9/57 20/9/57 4/9/57 20/9/57 t t
Chemical index Illillimo]es per Kg/mole ratio )
Chloride 5·8 7·8 3 19 27 20 54 /chloride-ammonium ratio /45 /1010 /820 Z /chloride-fluoride ratio /i 20 150 /430 155 0 /chloride-sulphate ratio /1D /65 /130 157 /224 /22 123 /chloride-boric acid ratio /i·2 /24,5 /24 /34 /chloride-arsenic ratio ,'19 th /30 /21·5 /1260 /720 Sodium 20 20 8 23 34 20 49·5 /sodium-potassium ratio /12·5 /29 /37 124 111 /40 /12·1 /sodium-lithium ratio 170 l6i 116 /40 /sodium-calcium ratio 113 135 /42·5 /930 /115 /59 Calcium . . 0·05 0 ·15 0·9 /calcium-magnesium ratio /0·6 l15 17 Sulphate 0,4 0·9 /sulphate-ammonium ratio 14·1 /50 f/9 13
*See note at end of Table 7c. tMean ratio from the analyses of two samples taken at different times. rABLE 108: Chemical Indices* for Analyses Given in Table 9B ( Analyses Made in 1958 and 1959)
Wairakei, Holes§ 6 6 7 Wairakei, Holest Hole No. 4 ( For Comparison) ( For Comparison )
23/1/59 10/7/59 Date Sampled 23/1/59 22/9/58 and 15/10/58t
Chemical Index ( millimoles per Kg/mole ratio) 63·0 0-2 41 35·5 60·3 1 0·2 Chloride . . 55·6 31 /4800 1 300 /270 /4100 1 300 /chloride-ammonium ratio 191 /165 2 4 /162 1 3 /chloride-fluoride ratio /190 /87 /172 1 7 136 t15 /178 1 5 /chloride-sulphate ratio /31·5 /25·5 /24·0 -1- 02 /24·1 1 0·2 /chloride-boric acid ratio /ji /31 /890 /995 1 40 /965 1 40 /chloride-arsenic ratio /980 38 26 57·1 : 0·2 55·8 _1- 0·3 Sodium 51 28·5 /\5 /12·7 2 0·4 19·8 30·2 /8·9 /10·5 19-2 /27·0 -1 0·1 /sodium-potassium ratio 139 /38 /29·1 1 0·6 /sodium-lithium ratio , . /38 /38 /200 1 7 /105 136 /114 z 17 /sodium-calcium ratio /122 0·31 1 0·02 0·4 0·9 0·49 .5 0·02 Calcium 0·3 to·6 /4'·9 i 0-8 /3·2 1 0·2 /calcium-magnesium ratio /2·3 11-3 0·31 1 0·02 0·37 0·02 Sulphate 0·85 /23 4 2 /28 4 2 /sulphate-ammonium ratio /70
tMean of values for two samples. iHoles drawing from "permeable volcanics", weighted mean values with *See note at end of Table 7c. §Holes drawing from fissures in ignimbrite, weighted mean values with standard deviations. standard deviations. TABLE 11: Analyses of Steam Separated from the Discharges of Holes at Waiotapu
Hole No. .. 3 6* 7
Date of Sampling 4l9f57 22/9/58 and 10/7/59 15/10/58
Wellhead pressure lb/sq. in. gauge...... Sampling pressure lb/sq. in. gauge 66 108 50 8 30 20 Enthalpy of discharge by gas content determinations at iwo pressures Rumb .. 323 Enthalpy of discharge Btu/lb (M.O.W. measurements) . . Steam content of discharge at 0 lb/sq. in. gauge %.. 334 430 450 Steam content of discharge at sampling pressure X 14·.9 26·0 28.0 * Gas content of steam at sampling pressure, %by volume,...... 12-7 20-2 23,8 w Calculated gas content at 0 lb/sq. in. gauge, %by volume.... 0·185 0·162 0·420 Calculated gas content at 0 lb/sq. in. gauge, %by weight..... 0·157 0·125 0·357 0'37 Calculated gas content on total discharge (free CO 2 and H :S in water phase negligible) 0·31 1·07 moles per 100 moles 0·023 0'033 0·100 Analysis of steam separated at sampling pressure Ammonia millimoles/100 moles....,. Analysis of gas in steam (analysis' on air-free samples) 0,245 0,70 Carbon dioxide %...... ,... T Hydrogen % .. .. 91·00 87·3 92-,2 1·93 Methane % . . 1·0 2·0 Ethane % 0·47 0·2 05 Nitrogen % .. 0·015 Hydrogen sulphide. %.. .. '; • 1·95 0·4 Ammonia % 4·50 10·7 5.3 0·13 0·44 n.d.
*Mean of values for samples taken at different times. tGas analysis on sample taken 22 September 1958. Hole 7: The discharge of this hole is very similar to that from Hole 6. The main difference is that the gas content as indicated by the low Cl/CO 2 ratio (table 12) is much higher. The chloride/total-sulphur ratio is, however, the same.
THEBEARING OF THE FINDINGS ON THE INVESTIGATION OF THERMAL AREAS During the chemical investigations at, Wairakei it had become apparent to the author that the study of a thermal area with a view to predicting its value for the production of power was rather unsatisfactory if only samples from the natural activity were available. There is first the difficulty that unless the springs are very hot, (i.e; well above boiling point deep down in the spring) and have very large. flows, they are liable to be affected by accidental local circumstances, such as taking up constituents carried by steam, or losing others by absorption iii the ground. Further, it is not possible to determine the gas content of the "primary chloride, water". nor the ratios of chloride to carbon dioxide, to total sulphur, or to total ammonium. It seemed, however, that all that was required to sample the area satisfactorily was a few prospecting holes, sunk just deep enough to reach the chloride water. lt is true that experience at Wairakei has shown that in many cases the early discharge is low in chloride content. presumably due to some dilution with cold water at the upper surface of the chloride water. At the same time there may also be heating by steam to give an enthalpy as high as that of deeper water..It may be necessary for the hole to discharge as long as a year before a constant value for the chloride content is obtained. The ratios of chloride to other constitu- ents can be obtained from early samples of the discharge, and even the ratios of chloride to those constituents that occur in water and in steam are not much in error. it seems, however, that the problem is more complex. Samples from six holes have been obtained, and a complete analysis ofthe most important 66 primary chloride water" at Waiotapu is still not available. It is suggested that from two to four distinct bodies of chloride water may exist. If it is possible for a "primary chloride water" of high chloride content to exist at depth without giving a spring like Champagne Pool, or any surface flow, then an area like Waiotapu could be sampled by a few holes, and perhaps rejected as a source of power when in fact there was a good high-chloride source of heat at depth. This problem did not arise at Wairakei, where any one of the prospecting bores, apart from those that struck bicarbonate water, gave a fairly good sample of the "primary chloride water".
AN HYPOTHESIS TO EXPLAIN THE CHEMICAL FINDINGS AT WAIOTAPU The fact that the discharges from Holes 3 and 5 agreed better in com- position with the water of Postmistress Pool was thought at first to indicate that there was a body of water of low chloride content possibly
113
5 in the form of a lens, above the higher chloride water which supplied Champagne Pool. Hence Hole 3 if deepened might penetrate into water of higher chloride content, especially as there was a suggestion that the bottom temperature was increasing after having remained steady for some time. However, when the results of investigations by Lloyd ( 1959 and Chapter V) became available and map 2 showing the geology and possible faults had been drawn, a different conclusion was possible. It appeared that different types of water were associated with the various faults. The high chloride water seemed to be connected with Champagne fault. Water similar to that from Postmistress Pool is associated with the Opouri and Kakaramea faults. There may also be a different type of water associated with the Waiotapu fault running past Lady Knox Geyser (N85/6/52). The Ngapouri fault is associated with springs very low in chloride, although the water from Hole 1 had a definite content of chloride.
lf this deduction is correct it appears likely that the thermal activity at Waiotapu is at an earlier stage than that at Wairakei, at least on the supposition that the heat source at Wairakei is directly below the reservoir of uniform chloride water. On this supposition, the difference at Waiotapu would be that there had not been time for the mixing of the difTerent waters associated with the separate faults. However, from the work on sulphur isotopes, as well as from the ammonia and nitrogen content of the underground water at Wairakei ( Rafter, Wilson, and Shilton, 1958), the author has come to the conclusion that Wairakei must be supplied laterally from a source at some distance. By contrast, the source of the magmatic steam at Waiotapu is thought to be directly below the Waiotapu area, so that it has not been possible for mixing to give a uniform body of water. The hypothesis is put forward that at the beginning of the activity magmatic steam comes to the surface along various faults, but that hydrogen chloride originally in the steam is held in the ground at depth. With time, ground water penetrates to the chloride-containing strata, and "primary chloride water" is formed. This water may still be heated by magmatic steam, but part of the heat will be supplied by metamorphic reactions in the rock and part by conduction of heat, including heat of solidification of the magma. The varying ratios of chloride to fluoride and boric acid will depend on how long the chloride has accumulated and on how much fluoride and boric acid is still supplied by magmatic steam. Hence there is the possibility that different ratios will be associated with different faults.
It is known from the work of Grimmett and Mcintosh (1939) that there is much arsenic at Reporoa. They give no chloride/arsenic ratios but the present work shows that the ratios increase progressively from Lake Ngakoro to Postmistress Pool. From work by the author (Wilson, 1959, p. 40) at White Island, arsenic seems to be held at depth even more than chloride. It seems possible, therefore, that the activity is oldest in the most southerly part of the area, where more chloride and arsenic have accumulated. Development of thermal activity may have occurred more recently in each successive fault to the north.
114 TABLE 12: Chemical Indices* Calculated on Total Discharge (i.e., on the Underground Chloride Water) for Constituents Present in Both Steam and Water, or Present in Different Chemical Combinations
Hae No. 3 6t 7 Wairakei Holes If Wairakei Holes § for Comparison ' for Comparison
Date of sampling 419151 22/9/58 and 10/7/59 15/10/58
Chen,Ical Index ( millinioles per kg water/niole ratio) 45·1 0·6 Chloride 16·6 22·8 32·2 43·2 1 0·5 - 1 --A /chloride-ammonium ratio , . 7440 /200 /510 1 70 /595 1 45 /chloride-total carbon dioxide ratio /1·3 /1 ·32 /0·47 /5·2 +0·7 /4·30 1: 003 /chloride-total sulphur ratio /10·6 /8·4 8·5 /66 17 /61 1 2 0·74 j 0·02 Total sulphur 16 2·7 3,8 0·60 . 0·06 /total sulphur-sulphur as sulphate ratio /2·5 11.4 22·7 1 0·3 12·82 1 0·14
Sulphur as sulphate.... 0·75 () · 62 0·22 1 0-01 0·26 i 0·01 /sulphate-ammonium ratio.... 120 /5·3 /2·9 i 0·4 /3·5 i 0·04
*See note al end of Table 7c. tMean of values for samples taken on 22 September 1958 and 15 October 1958. iHoles drawing from "permeable volcanics", weighted mean values, with standard deviations. §Holes drawing from fissures in ignimbrite, weighted mean values with standard deviations. CONCLUSIONS The chemical work has shown that there is a natural discharge of hot water at Waiotapu which is very similar to the hot water found beneath the Wairakei area. However, none of the holes drilled so far at Waiotapu has tapped water of similar composition to this natural flow (with the doubtful exception of Hole 4). One possibility is that the holes have to discharge first of all a large amount of altered water which overlies a body of water of uniform composition. This was found to be the case at some but not all the holes at Wairakei. This does not seem to be the likely explanation at Waiotapu, for the differences are very much greater. Not only is the chloride lower, a circumstance simply explained by dilution with cold surface water, but also the chemical indices are markedly different. There would be no simple explanation as to how water discharged by Hole 3 could be derived from water of composition similar to that issuing from Champagne Pool. From the experience at Wairakei it would appear that successful development for power of the Waiotapu field could be expected if the bores could tap water similar to that discharged by Champagne Pool. From the large flow of this spring it can be argued that there must be underground a large amount of water of high chloride content. The chloride content can easily fall as hot water comes to the surface, but apart from a small concentration due to loss of steam there is no way by which the chloride content can be raised considerably. The first objective in further prospecting of the field should be to locate bores to tap this high chloride water, and it appears that more bores are required to the south of those already drilled. From the considerations in the previous section it seems, however, that there is quite a possibility that the size of the body of water of high chloride content may be quite limited, and that in contrast to Wairakei there is no extension over a large area of hot chloride water of uniform composition. On the other hand there is the possibility that sources of magmatic steam are directly beneath the area, so that unsaturated steam, free from water, might be tapped by drilling. Such steam would be much better for power generation that the wet steam obtained at Wairakei. However the hole to tap such steam would need to be much deeper and would be more difficult to site. It is obvious that holes not correctly sited may be useless. At Wairakei, any hole drilled in the area under which the uniform body of hot water is known to occur is likely to give a fair yield, and at the same time still better results are to be expected as the feeding zones are located. Present indications are that the difficulties of locating good holes at Waiotapu may be much greater, but that the final results may be better than at Wairakei. Further chemical work could be directed to obtaining additional evidence for or against the hypothesis put forward above, that waters of different composition are associated with the various faults crossing the area and that the activity is more recent towards the north. The best evidence would be obtained from the discharges of holes drilled at inter- vals along the faults. Even shallow bores that did not discharge could be
116 quite useful if they were sampled with an improved sampling bottle. It is unfortunate that samples from Holes 1,4, and 5 were not fully analysed. However, drilling is costly, and it may be that results so far have not been good enough to encourage the continuation of this work. In that case, one is dependent. on samples from natural activity. There does not seem to be much hope of obtaining information from spring waters. The springs not yet sampled are too small, too low in chloride, and too acid to give useful information, and indications are that they are too much affected by purely local circumstances. As gas seems plentiful throughout the area, a more promising approach might to be collect it systematically from both fumaroles and bubbling pools. Determinations of the gas content of the steam are desirable at the fumarole area on Maungaongaonga, and, if possible, at the acid-sulphate area further to the north-east along the Ngapouri fault (springs N85/3/16-20) in order to ascertain whether the steam is of direct magmatic origin. As regards the composition of the gas from bubbling pools, it appears that there are only two useful figures from the gas analyses, the ratios of carbon dioxide to hydrogen and to methane, and of these two, the utility of the methane figure is somewhat doubtful, owing to the possi- bility of some of the methane having its origin in decomposing organic matter. Other determinations are required, and the only possible ones seem to be the ratios of stable isotopes, particularly those of carbon and sulphur. The difficulty with determinations on sulphur is that sulphur occurs in the forms of both sulphate and hydrogen sulphide, and the latter may be oxidised at the surface to give elemental sulphur. In a paper by Rafter, Wilson, and Shilton (1958) it was shown that a figure for the temperature at which sulphate and hydrogen sulphide had been in equilibrium could be deduced from the stable isotope measurements on samples from holes. The determination of this temper- ature from samples of sulphate from waters and hydrogen sulphide from gas might be useful at Waiotapu. At any rate, it seems that chemical and physical work on gas samples is the work most likely to throw light on the problems raised in the present paper.
ACKNOWLEDGMENTS The chemical analyses of the hot-spring samples collected iii 1955 were made by Messrs A. J. Ellis, G. A. Patchett, J. F. Thompson, 1. C. Cloke, and Miss A. C. Camden-Cooke in the Geothermal Chemistry Section of the Dominion Laboratory, Wellington. The samples collected by the author in 1937 were analysed by Mr J. A. D. Nash. Some of the analytical work on the samples of bore water collected in 1957 was done by Mr A. J. Hilton at the branch of the Dominion Laboratory at Wairakei. and the remainder by Mr C. K. Backhouse in Wellington under the direction of the author. Some samples from hot springs and bores were collected by Mr W. A. J. Mahon in 1958-9 and analysed in the laboratory at Wairakei. An analysis of the gas was made by Miss J. Ross, of the Coal Section. Dominion Laboratory.
117 REFERENCES ELLIS, A. J. 1961 ; Geothermal Drillholes - Chemical Investigations. Working paper for U.N. Coid. 0,1 New Sources of Energy. E/Conf. 35/G/42. ELLIS, A. J.·; WILSON, S. H. 1955: The Heat from the Wairakei-Taupo Thermal Region Calculated from the Chloride Output. N.Z. J. Sci. Tech. B.36: 622-31. GRANGE, L. I. 1937: The Geology of the Rotorua-Taupo Subdivision. N.Z. geo/. Sm·v. Btill. 31 (n.s.) GRIMMETT, R. E. R.; MCINTOSH, I. G. 1939: Occurrence of Arsenic in Soils and Waters in the Waiotapu Valley, and its Relation to Stock Health. N.Z J. Sci. Tecll. A2l . 137-45, LLoYD, E. F. 1959: The Hot Springs and Hydrothermal Eruptions of Waiotapu. N.Z. J. Geol. Geophys. 2; 14\-16. RAFTER, T. A.; WILSON, S. H.; SHILTON, W. B. 1958: Sulphur Isotopic Variations in Nature Part 5 - Sulphur Isotopic Variations in New Zealand Geothermal Bore Waters. N.Z. J. Sci. 1. 103-26. WILSON, S. H. 1955 : "Chemical Investigations". /n Geothermal Steam for Power in New Zealand. N.Z. Dep. sci. indi,str. Res. Billi. 117·,pp. 27-42. - 1959: Physical and Chemical Investigations. hi White Island. lbid. Bull. 127: Pp. 32-50.
118