Proceedings 20th Geothermal Workshop 1998 NGAWHA GEOTHERMAL FELD A REVEW M.J. School of Mathematical Computing Sciences, Victoria University Wellington, NZ SUMMARY- The Ngawha geothermalfield is different to other New Zealand fields, in its location, geology and pressure responses. Reservoirproperties relevant to modelling are reviewed, including pressures, temperatures and pressure transient modelling. The potential for heat-sweep with production-injection doublets is upwards of over30 years. A developmentof 2-3 has the possibility of negligible impact on surface features at Ngawha. 1. NTRODUCTON This is a review of what is known about the Ngawha geothermal field, from the perspective of reservoir engineering. The aim of this review is to focus on the potential for development at Ngawha. t is a personal selection the wealth of information publicly available. Major sources of material used are Mongillo (1985) and Cox (1985). Ngawha is unique among the known hightemperature geothermalfields in New Zealand, in being located outside the Taupo Volcanic Zone, to the north of Auckland(Fig. 1). Permeability is associated with faulting in the basement greywacke formation, at depths of 1600 approximately. This is overlain by 500 m of formations that effectively cap the reservoir. 60 Figure 1. Ngawha and the Taupo Volcanic Zone Locations (Mongillo, 1985) 411
Fluid is interpreted to rise in the north and southeast, high in gas (mainly and partly degas and flow laterally to the south and south-west. The north and east of the field are not delineated by drilling, whereas the southern and western boundaries are. There is significant recirculation, and entrainment of cooler groundwatersfkom the east. Some fluid, about makes its way to the surface through the caprock, driven by a 10-baroverpressure (Grant, 1985). The recirculating reservoir fluid is typically high in boron and ammonia, with similar NaCl concentrationsto other fields. levels are about twice those at Ohaaki,and ten times those at Wairakei. 1.1 Surface Features There are a number of hot springsand baths in the Ngawha area, traditionally valued for their healing powers (Dieffenbach, 1843). Some of these are shown in Fig. 2 fiom (Sheppard and Johnston, 1985). Natural discharge is low in fluid at less than 2 but with high gas flows and with high boron content in the fluid, compared to other New Zealand geothermal fields (Sheppard and Johnston, 1985).Hot gassyupflow throughthe low permeability caprock mixes with local groundwaters to create a fluid that is high in bicarbonates and magnesium and chloride. e-- -\ f a 1000 1200. Figure 3. Reservoirpressures versus depth Laterally, NG2, NG3 and NG8 appear to be a bar or two higher in pressure than the overall linear trend. Grant (1985)has used this to infer a horizontal permeability-depthof 70 darcy-metres, assuming that there is a 1 lateralflow of reservoir fluid. Subsequent analysis of interference tests suggeststhat horizontal permeability through the system of fractures that defines the reservoir is significantlyhigher than the value used by Grant. This suggests a much higher inferred lateral flow of reservoir fluid, perhaps a factor of ten or one higher. West of field NNW Ngl Ng9 Figure 2. Natural Thermal Features at Ngawha (Sheppardand Johnston, 1985) 2. PRESSURES The interpretedreservoir pressures are reproduced against depth from Grant, 1985,in Fig. 3. The vertical pressure gradient indicates downflow overmuch of thefield, as it isbelow hydrostatic for fluid. This is consistent with models that postulate much of the permeable reservoir is outflow of deep hot sourcefluid entrained with cooler meteoric waters. Figure 4. The West Profile of interpreted temperature contours with depth, fiom Cox (1985). The heat and mass balance of McNabb and Mc- Donald (1982) leads to an inferred 50-100 inflow of deep hot source fluid into the reservoir. Their model includes the entrainment of cool groundwater, and uses chloride and heat balances to obtain flowrates. Such entrainment is also suggested by temperature inversions and dilution trends across the reservoir. Hence the total lateral flow will be greater than 100 as the deep hot upflow is augmented by entrainment 412
of cooler groundwater. ndeed, the entrainment model of McNabb and McDonald (1982) has the total outflow rate increasing exponentially,so that at a 5 radius the radial outflow would be as high as 400 in their model. However, their model does not allow for a regional throughflowof groundwater,and must break down at some radius from the upflow. Determination of the regional throughflow rate remains an outstanding problem at Ngawha. 3. TEMPERATURES A table from Grant and McGuinness, 1985,of values of interpreted temperatures (and pressures) is presented in the appendix. Figs. 4 and 5 (taken from Cox, 1985) use these values, and present interpreted temperature contours in two vertical slices through the reservoir. The slices are the West Profile and the East drawn as heavy lines in Fig. 6. Also shown in this figure from Cox (1985) are the basement faults and drilled wells. Fig. 7 (from Cox, 1985) presents the interpreted temperature at the interfacebetween reservoir and caprock. The dashed lines indicate the three regions of surface thermal activity. n the vertical sections, fluid flow directions have been interpreted and indicated by arrows. An interestingdetail in these figures is the inferred (minor) upflow at the Mangatawai fault in the southeast of the field. Although the temperature measurements suggesting this upflow are sparse, there is also evidence for the upflow in the chemistry of dischargedfluids (Cox, 1985). Figure 6. Plan view showing the West and East Profiles (Cox, along which Figs. 4 and 5 are taken. Pressure responses are controlledby fluid recharge from the lower permeability rock between fractures, and from the boundaries of the reservoir. Fractured reservoir models (Grant and ness, 1985, and McGuinness, 1986) give a reservoir volume of 70 15 and a fluid volume of 2.6 0.6 with 20% of fluid residing in fractures.. Figure 7. Temperatures at the interface between caprock and reservoir, from Cox (1985). Figure 5. The East Profile of interpreted temperature contourswith depth, from Cox (1985). 50 4. PRESSURE TRANSENT TESTS 40 The conventional homogeneous reservoir models are remarkably inadequate when fitted to data from longer term interference tests at Ngawha (McGuinness, 1986). The most successful models are dual porosity (fracture-block)with practically infinite permeability in the fractures. Fig. 8 shows data (boxes) from a one-month interference test in 1983, with fitted Theis solution (homogeneous reservoir model-thinner solid curve) and a fitted bounded randomly fractured reservoir model (thicker curve). The fractured reservoir model fit is almost indistinguishable from the data. 30 20 - - data fracture -Theis solution Figure 8. Pressure response in models (McGuinness, 1986). and fitted 413
5. EXTRQCTNG ENERGY FROM THE RESERVOR n the early strategies for developing Ngawha centred on conventional separation of steam and use of downdraft steam turbines to generate electricity. Such a development would be hampered by the large amounts of ble gas present in the reservoir fluid, and in the absence of reinjection would also lead to boiling and possibly fallingpressures in the field. mpact on the hot springs and mud pools would have been inevitable eventually under such a development, although surface features might temporarily have been boosted by extra steam produced by boiling in the reservoir. However, reinjection of all produced fluids (and possibly gasses) would reduce the pressure impact on the field, and would suggesta heat sweep model of the reservoir,one that took intoaccountthe tured nature of the greywacke. Small scale development even has the possibility of not impacting on the surface features, since the outflow can sweep away cooler reinjectedfluids before they return to the production region. Figure 9. A sketch indicating the permeable region drilled at Ngawha. Also shown is an approximate 220 C contour in the reservoir, and the resistivity boundary. 5.1 Heat-Sweep Models f most produced fluid is reinjected into the permeable reservoir, and pressure changes can be neglected, then temperature changes may by modelled by considering the heat swept out of the reservoir by a production-injection doublet. Possibly important modelling features include a homogeneous reservoir, a double-porosity reservoir (reflecting the important role played by fractures when consideringfluid flow at Ngawha), a reservoir with regional and a reservoir with deep hot upflow. The amount of stored heat in place at Ngawha within the permeable drilled region sketched in Fig. 9 may be calculated. Assuming the reservoir has a vertical extent of 700m and an areaof 3 that the volumetric heat capacity of rock and fluid is 2.5 x "C,and that a 60 C temperature difference exists between produced and reinjected fluids, the available thermal energy is 3 x J. This is equivalent to about 30 for 30 years. alternativeestimate of reservoir volume comes from the analysis of interference tests ness, which imply a total reservoir volume of 70 10 This leads to the larger estimate of J of thermal energyavailable,equivalentto about 1000 for 30 years. This would be an over-estimate,since 1. pressure communication can be expected to extend into cooler outer regions of the reservoir, and 2. the interference test analysis used the compressibility of water assuming a sealed reservoir, whereas two-phase fluid is known to be present, and the reservoir might not be responding as a box (there may be water level changes with some communication with the atmosphere). Both effects increase the apparent compressibilityby factors of 10 or more (Grant et and would decrease the estimated available energy by the,,same factors. The above calculations of heat in place ignore the effects of natural thermal input, estimated by Nabb and McDonald at 50- f 300 of this is available,this gives about 2-3 More sophisticatedmodelling, allowing for the effects of on thermal returns to production wells (Bodvarsson and Tsang, 1982) or allowing for two-dimensional heat sweep by fluid streamlines and including the effect of a regional throughflow(gringarten and Sauty, is possible. This modelling does not substantially alter the heat-in-place results above, but does provide more detailed predictions of temperature changes at specific production wells, and of the size and shape of cooled regions near injection wells. 6. CONCLUSONS A number of wells suitable for production and injection are already in place at Ngawha. Any development will have to cope with high levels of condensible gases. The permeable drilled reservoir that is presently known is an outflow a downflowregion-the deep hot source is not accuratelylocated yet, although there is circumstantial evidence for upflow in the north and and the 414
south-east. A developmentof 2-3 in existing wells would balance the estimated natural hot inflow. Smaller developments have the potential of little or no impact on existing surface features at Ngawha, depending on well locations. 7. REFERENCES Bodvarsson, G.S. and Tsang, C.F. (1982) njection and Thermal Breakthrough in Fractured Geothermal Reservoirs, Geophys.Res. 1031-1048. Cox, M.E. (1985) Geochemical examination of the active hydrothermal system at Ngawha, Northland, New Zealand: model, element distribution and geological setting, being a thesis submitted to the GeologyDepartment, University of Auckland. Dieffenbach, E. (1843) Travels in New Zealand (2 volumes)., Reprinted in 1974, Copper Press, Christchurch,NZ, pp. 243-255. Grant, M.A., Donaldson,.G. and Bixley, (1982) Geothermal Reservoir Engineering, Academic Press, New York. Grant, M.A. and McGuinness, M.J. (1985) Hydrology, Chapter 13 of The Ngawha Geothermal Field: New and Updated Scientific nvestigations, DSR GeothermalReport Number 8, compiled and Gringarten,A.C. and Sauty, J.P. (1975)A Theoretical Study of Heat Extraction From Aquifers With Uniform Regional Flow, Geophys.Res. pp. McGuinness, M.J. (1986) Pressure Transmission in a Bounded Randomly Fractured Reservoir of Single-phase Fluid, Transport in Porous Media 1 McNabb, A. and McDonald, W.J.P. (1982) An energy balance model for the Ngawha geothermal field, N. Z.Journal Science pp. 167-173. Mongillo, M.A. (1985) The Ngawha Geothermal Field: New and Updated Scientific nvestigations, DSR GeothermalReport Number 8, compiled and Sheppard,D.S. and Johnston, J.R.(1985)Monitoring of thenatural Thermal Featuresof the Ngawha Springs, Chapter 8 of The Ngawha Geothermal Field New and Updated Scientific nvestigations, DSR GeothermalReport Number 8, compiled and APPENDX Table of well data values, from Grant and McGuinness (1985). N63 M3 DATE DRLLED 12.64 213 03.81 213 02.79 563 61 1467 08.78 04.79 210 116 253 274 652 626 626 1189 06.82 230 954 626 07.82 22 16% ELEVATON 579 535 740 700 670 950-30s -1037-740 -697-366 222 230 226 226 224 226 59.2 110.2 71.6 94.S 92.5 92.3 63-0.4 10-2 0.9 9-30 110 52s 910 955 985 955 91 100 220 945-1129 230 24 1 L NG2 was drilled to a depth of on 11/12/77,then deepened to on 11/04/78. * NG2 was apparently improvedby a stimulationtest in 198 P = Primary feed zone. S = Secondaryfeed zone. 415