Wairakei – A Fresh Look at Rejected Technologies

BRIAN R. WHITE

GRANT L. MORRIS

CHRISTOPHER TAYLOR

Full addresses/phone/fax

Genzl division of PB Merz and McLellan (NZ) Ltd, P.O. Box 668, Wellington, N.Z.Ph. (04) 499 1000 Fax. (04) 495 8639

ABSTRACT - 1998 marks the 40th anniversary of the commissioning of the first units at WairakeiGeothermal Power Station, New Zealand. While precedents for geothermal power development hadalready been set in Italy with the harnessing of vapour-dominated fields, and in Japan on a small scale,it was Wairakei that led the way for large-scale development of water-dominated fields for powergeneration. At this 40th anniversary it is interesting to look back at some technologies that wereconsidered for Wairakei and rejected, either prior to commissioning or shortly after, and then to seewhere these rejected technologies have gone since. Amongst these were binary cycle technology, aheavy water distillation plant, mechanical gas extraction, centralised multiflash systems and simplemethods for phase separation.

1. BINARY CYCLE GENERATION

In the late 1940’s and early 1950’s water-dominated steamfields had not been developedcommercially for power generation. There wasan expectation that fluid conditions could bemore arduous than those found at Lardarello.(Later materials studies showed that Wairakeifluids were quite benign). New Zealandengineers had studied the facilities at Lardarellobut that technology could not be directly appliedto New Zealand fields. Either phases had to beseparated to allow steam to be used, or analternative technology had to be developed thatcould handle wet steam.

One solution trialled at Lardarello for handlingthe difficult fluid conditions (particularly thehigh gas content) was to install heat exchangersfor the generation of clean steam. However,right at the outset of consideration of powergeneration at Wairakei, the question was askedwhether a water-based working fluid was themost appropriate to use. The New ZealandDominion Laboratory was commissioned toinvestigate the use of Organic Rankine or BinaryCycle technology with a view to direct use of 2-phase fluids.

The brief reports on this concept have provedelusive, but minutes of a geothermal meetinginvolving staff from the Department ofScientific and Industrial Research and from theMinistry of Works in September 1950 makereference to the “ethyl chloride” plant as analternative generation means. It is thought thatthe study did not proceed far beyond aconsideration of a working fluid and the

associated data, supply and cost problems.However, it is significant that a geothermalapplication for this technology was beingconsidered at such an early date. At that time,the Organic Rankine Cycle was almost unheardof.

There had been some early consideration oforganic fluids as working fluids for hightemperature applications, but research paperstended to discount such use (Alcock et al, 1953).Bronicki (1984) records that development ofOrganic Rankine Cycle turbines began in thelate 1950’s with a view to solar applications(Tabor, 1961). Another research paper(Aronson, 1961) records the existence of vapourturbines at that time, while suggesting theapparently novel use of organic fluids (renamedrefrigerants) in power cycles. However, aslightly later paper (Luchter, 1967) reviewed thestatus of technology, referencing somedemonstration plants and concluded that thetechnology could seriously be considered for“near term application”. Hence, it appears thattechnology to implement the Wairakei binaryconcept did not exist for another one or twodecades after the initial concept.

The first known implementation of binarytechnology for a geothermal application was inthe USSR at Paratunka in 1967 (output was lessthan 0.5MW), nearly two decades after theWairakei binary concept (DiPippo, 1980). Chinafollowed with a number of plants of small scalein 1970 (Zhou, 1983). The plant at Wentang,commissioned in 1971 is notable for also usingethyl chloride (or R160) as the working fluid.

Acceptance of binary technology was slow, withthe focus in the geothermal world being oncontinued use of high temperature fields andflash technology applications. However, generalinterest was accelerated by the energy crisis inthe 1970’s, which significantly altered energycosts. Developments occurred in the areas ofsolar power, geothermal power and waste heatrecovery (Angelino et al, 1984). A rapid boostcame with the establishment in the US ofrenewable energy incentives and theestablishment of Ormat Energy Systems Inc inthe early 1980’s.

Amongst a number of designers andmanufacturers, Ben Holt and Ormat became thedominant players in the geothermal binary cyclemarket. Ben Holt designs were based on super-critical cycles using iso-butane as a workingfluid while Ormat designs are based on sub-critical cycles using pentane as the workingfluid.

Ormat have continued to expand applications ofbinary cycle technologies with the use of hybridsystems (Figure 1). These use phase separationwith the steam phase being passed through asteam turbine to almost atmospheric pressure.Binary cycle units then replace the condenser.The liquid phase is usually passed throughdedicated binary units to generate additionalpower. These hybrid plants have similar costsand performances to conventional plants. Thefirst was installed in 1992 at Puna, Hawaii. Thelargest has been installed at Upper Mahiao,Leyte, Philippines. In New Zealand, this hybriddesign has been used at the recentlycommissioned Rotokawa plant and is being usedin the design and construction of the first stageof the Mokai plant.

Figure 1: Typical Hybrid Binary CycleSchematic

Contact Energy Ltd has recently applied forconsents for a simple binary plant (not thehybrid type) to be installed at Wairakei to takeadvantage of the hot fluid now being piped to

the vicinity of the existing power station forreinjection purposes.

Binary cycle technology has moved from vagueidea to established technology since the initialWairakei concept. It is now routinely consideredas a geothermal power generation option forlower temperature resources, and in its hybridform for high temperature resources. Itsacceptance in connection with geothermalapplications appears to be better than for non-geothermal applications, where limited use isfound for waste heat recovery.

2. HEAVY WATER DISTILLATIONPLANT

The New Zealand government was committed toa Wairakei development from the early 1950’s,with Wairakei firmly established as a projecteffective from 30 July 1952 (Bolton, 1998).However, the government’s resolve to developthe Wairakei geothermal field was helped by theinterest of the United Kingdom Atomic EnergyAuthority (UKAEA) in the production of heavywater for it's nuclear programme. New ZealandGovernment approval in principle was given inMay 1953. In 1954, the Government formallyannounced that a joint venture had beenestablished between the New ZealandGovernment and the UKAEA to develop a plantproducing 40MW of electric power and 6t/yearof heavy water. The British firm of Merz andMcLellan was appointed to undertake the designof the power station while Head Wrightson wasappointed to design the heavy water productionfacility (Love and Bolton, 1980).

Details of the heavy water plant are not known.However, the Wairakei location was chosenpartly because of the availability of largequantities of inexpensive heat. The process oflarge-scale production of heavy water at thattime required much hydrogen gas andelectricity. A hydrogen-water exchange reactionwas employed in which hydrogen gas and steamwere passed upward through a tower containinga catalyst, while liquid water flowed downward.An isotopic exchange reaction took placeresulting in deuterium concentrating in the steamwhich then condensed, and was passed out ofthe bottom of the tower with the rest of thewater. Consequently, this water was enriched inD2O. A cascade arrangement lead to furtherconcentration, with the final concentration beingachieved by electrolysis (Glasstone, 1967).

It may be significant that all correspondencerelated to the heavy water plant refers to the“distillation plant” only. There is no mention of

hydrogen or electrolysis. A demonstration plantand corrosion test facility was established onWK9. This was essentially just a distillationplant incorporating corrosion coupons within thedesign. The plant was apparently successful indemonstrating the principle of distillation andshowed that ordinary materials such as mildsteel were adequate for geothermal applications.Effort went into ensuring that the corrosion datawas retained no matter what the future of theheavy water plant was.

The pioneering nature of the designs presentedproblems, and there are some verbal reports thatsimple arithmetical errors affected the initialconcept. The initial design of the powerstation/heavy water plant required twoseparation pressures in the field, one highpressure system at 13.5barg and a secondintermediate pressure system at 5.5barg. This inturn was associated with station pressures of12barg and 3.5barg. The high pressure was forelectricity generation while the second pressurewas dedicated to the heavy water plant (Bolton,1977). The heavy water plant was designed totake 450t/h of steam. It was designed to produce6t/year of heavy water. Preliminary designconcepts developed in 1954 were refined in thedetailed design phase. By the end of 1955detailed engineering costings of the respectiveplants had doubled the cost of the heavy waterplant and increased by one-third the cost of thepower plant (although it’s output had increasedfrom 40 to 47MW). As a consequence, theUKAEA announced it’s decision not to proceedwith the heavy water project in January 1956(AJHR, 1956).

Design and procurement had reached anadvanced stage when the heavy water plant wasabandoned so station steam pressures wereretained. A decision was made to replace theheavy water plant with steam turbines. Thepressure drop and flow available lead to theselection of two 11MW sets. This enabled theuse of identical alternators to those of the LPsets. A comparison between the simple layout of“B” Station with the contrasting layout of “A”Station indicates the extent to which thisabandoned heavy water plant affected the designof the initial station.

The heavy-water distillation plant also dictatedthe LP manifold pressure as it was going toreject steam at 0.1barg. It was decided to installcondensing turbines to take this atmosphericsteam for additional generation. Theseatmospheric pressure turbines have successfullyoperated at Wairakei to the present withoutdifficulties. Atmospheric turbines represent an

opportunity for additional generation, whichcould be transferred to other stations or fieldsaround the world.

A later development for large scale productionof heavy water, possibly after the Wairakeitrials, utilised hydrogen sulphide gas andelectricity. As the first stage of production, thereis an isotopic exchange reaction between liquidwater and H2S gas. A “dual-temperatureprocess” is used whereby hydrogen sulphide actsas a carrier for deuterium giving it up to water atthe lower temperature (ambient) and regaining ita higher temperature (100oC). Depleted water isdiscarded while enriched water is passedthrough a cascade system to enrich it further.After a certain concentration is reached, thedual-temperature chemical exchange process isno longer economic, so the product is transferredto a distillation plant for further enrichment byfractional distillation. Finally, the production of99.75% pure heavy water (D2O) is achievedthrough electrolysis (Glasstone, 1967). If thepotential role of H2S gas in the production ofheavy water had been recognised at the time ofdesign, then it is possible that the Wairakeiheavy water plant could have proceeded.

Heavy water reactors for power production wereslow to develop, largely because of the cost ofproduction of heavy water. The UKAEA had itsown reactors at Harwell, but these were gascooled and graphite moderated. They were usedfor research and for space heating. The UKAEAwere probably spurred on in their powerresearch by the announcement from the NationalReactor Testing Station, Idaho, USA that theyhad generated the first electricity (60kW) fromnuclear power on 21 December 1951. Despitewanting to develop heavy water reactors, thecost and availability of heavy water forceddevelopment of gas-cooled reactors withgraphite as the moderator. Heavy water plantswere advanced in Canada, which had indigenousuranium resources but did not have thecapability to construct and operate a fuelenrichment plant. With a heavy water reactor, acritical assembly can be achieved using naturaluranium as the fuel. The first prototype CANDUpower reactor (the type developed in Canada)was commissioned in 1962 generating 22 MW(Glasstone, 1967). The UKAEA eventuallydeveloped its own version of the heavy waterreactor known as the Steam Generating HeavyWater Reactor (SGHWR). The only one of itstype in existence was the Winfrith 100MWprototype commissioned in 1967. Now the“favoured” reactor design is of the light watertype, and principally the Pressurised WaterReactor (PWR) which is derived from nuclear

plant for marine propulsion (Crossland etal,1992).

3. MULTI-PHASE TRANSMISSION ANDSEPARATION

Wairakei Power Station originally operated at 3different manifold pressures: HP of 12barg, IPof 3.5barg and LP of 0.1barg. As a result, onepossible misconception today might be that thesteamfield was originally based on multi-phaseseparation. This is not correct. Initially, wellswere tested then designated for use as HP or IPwells. The steam was separated at the wellhead,and separated water from all wells was feddirectly to the silencers and steamfield drains.

However, in the initial independent conceptsassociated with both the heavy water plant andthe power generation plant, the residual value ofthe hot water was recognised. The heavy waterproposal by Head Wrightson includedtransmission of hot water to the station forflashing and provision of additional steam forthe turbines. The power generation proposal byMerz and McLellan included the provision forflashing of additional steam at the wellhead.

At the time of power station commissioning in1958 a Pilot Hot Water Scheme wasimplemented and was commissioned by July1963, increasing power output by 5MW (Smith,1970 and 1973). This involved piping hot waterfrom the field to the station for secondary andtertiary separation (Figure 2). The designershowever were wary of steam formation and toprevent boiling in the pipeline, the high pressurewater was both pumped and attemperated with

Figure 2: Diagram of Pilot Hot Water Scheme

IP water. An elevated header tank in thesteamfield was used to maintain and controlpressure within the water line. Water wasflashed at the station to provide additional IPand LP steam. Effectively this was the first flashplant. The plant operated for several months butwas abandoned because of the characteristics ofthe wells connected to it. Field pressures weredropping rapidly in the first few years ofproduction reducing the water output of theconnected wells.

Another significant feature of the pilot hot waterscheme was the provision of scrubber vesselsafter each flash stage. Near pure condensate wasused to scrub solids from the water carryover toreduce the saline content of the IP and LP steamto less than 10ppm (Haldane and Armstead,1962). Although such scrubbers did not featurein subsequent flash steam developments atWairakei (the long steam pipelines effectivelyfulfilling that function), scrubber vessels havebecome an integral part of other geothermaldevelopments especially where steam separationis in close proximity to the power station.

The Pilot Hot Water Scheme was eventuallycannibalised. The hot water line was convertedto a steam line (H line). The separators werelater used in flash plants located in thesteamfield. Although the Pilot Hot Waterscheme had a short life, the feasibility ofsecondary or tertiary flashing had beenestablished.

A review of geothermal plants around the worldindicates a good proportion are dual flashsystems. Hence, although the technology wastemporarily abandoned at Wairakei, multi-flashsystems had been proven there and were later re-

implemented at Wairakei and many other fieldsworld-wide.

4. U-BEND AND T SEPARATORS

As part of the initial efforts to separate thegeothermal fluid phases, some very simpleseparators were tried. The New ZealandDominion Laboratory and the New ZealandMinistry of Works separately undertooklaboratory and field tests on the separationefficiency of simple Tees and U-bends withinside offtakes (Figure 3). As with the cycloneseparators that eventually replaced them, thesetook advantage of momentum principles. Theheavier liquid phase was less inclined to goaround sharp bends.

Figure 3: U-Bend Separator

The U-bend separator was originally used toseparate phases to conduct simple well tests. Itwas then applied upstream of the early cycloneseparators in order to prevent fragments of rockreaching the separator as well as assist withseparation. In such a role, the pressure dropthrough the U-bend was of more interest thanthe actual separation efficiency. And so, whilesome initial performance testing was done, therewas little development work done onunderstanding and improving efficiency. The U-bend “pre-separator” was soon made redundantwith the introduction of high efficiency bottomoutlet separators. The U-bend was quiteeffective and could be refined further with ourcurrent understanding of the two phase flowregimes within pipes.

The U-bend reportedly removed 80-90% ofwater, with the outlet steam about 50% dry.However, test results were quite variable. Thiswas thought to be due to the water leg operatingin a flooded condition. This meant that littlesteam passed down the water leg, and slugs ofwater had nowhere else to go than down thesteam leg. A larger vertical pipe or chamberwith water level control would avoid thisproblem today. In which case, separation

efficiencies greater than 90% could be expectedin the steam line. While this is not adequate forconventional developments, there are somesituations where this could be adequate.

Performance testing was conducted in the late1950’s when techniques for sampling andmeasurement were still being developed. At thetime, measurement error was recognised asbeing a limitation to understanding the separatorperformance.

Following preliminary tests at Wairakei, the U-bend separators were relegated to use as a pre-separator on some of the early wellheads.However it is worth considering theirperformance again, particularly with a view tobinary cycle or process heat applications. Ifsteam is required for a heat exchanger ratherthan a steam turbine, then the requirements fordryness can be relaxed. Steam dryness must behigh for admission to a steam turbine, both toreduce exhaust wetness and to reduce theconcentration of dissolved solids which couldeventually foul the machine. However, drynesscan be much less for a heat exchanger, with thedominant criteria being to avoid damaging flowregimes at the inlet to the exchanger. This canallow the relaxation of separator efficiencyrequirements by 3 or 4 orders of magnitude.Large separator vessels could be replaced withU-bend separators where flow regime allowswith a potential saving of millions of dollars forlarge developments. Consideration could also begiven to use of these in conjunction withdownstream demisters for steam turbine use.

5. MECHANICAL GAS EXTRACTORS

Wairakei steam, as with all geothermal steamcontains a certain amount of gas which has to beextracted in order to maintain condenservacuum. Plant in Italy utilised mechanical gasextractors of centrifugal design to handle gaspercentages up to 500 times greater than inconventional thermal plant. Wairakei’s initialconcentration was up to 50 times greater than forconventional plant (Haldane and Armstead,1962). Conventional thermal plant utilised steamjet ejectors to extract the gas principally sourcedfrom dissolved oxygen in the makeup water.

Mechanical gas extractors, with backup steamjet ejectors were initially selected for the “A”Station at Wairakei. Unfortunately the principleof relying on well-tried designs was difficult tofollow. The design was atypical, partly forcedby the low gas concentration and an apparentleaning by the designers towards the use of

mechanical exhausters, because of their commonuse in Italy. The Wairakei mechanical gasextractors operated at relatively high rpm(14,500rpm), as a consequence of the atypicalflowrates compared to the Italian conditions.Each LP condensing set had a 10-stage highspeed motor-driven exhauster in four separateseries casings. During commissioning, theseunits had problems with noise and excessivevibration (Bolton, 1977). Careful balancing andalignment was required. They proved veryunreliable mechanically, suffering from bearingand coupling failures. In addition, severaldesigns of motor rotor were tried out before adesign that could withstand the tough startingconditions was found (Haldane and Armstead,1962). They were eventually abandoned, withthe last unit taken out of service in 1970 (Staceyand Thain, 1983) and replaced with an all-steamjet ejector system still in use today (withmodifications). Although mechanical gasexhausters were considered for “B” Station, theywere ruled out at the tender stage forcommercial reasons (Haldane and Armstead,1962).

The design of the mechanical gas exhaustersselected for Wairakei is unusual by today’sstandards. The 10 stages and high speed implythat the pressure ratio achieved for each stagewas low, which may be explained by limitationsin technology, particularly the lack of inletinducers on the impellers, as well as the lowinlet flowrate.

Today’s mechanical gas exhausters are morereliable, simpler and run at slower speeds.Inducer impellers are used where the bladesextend down around the hub radius so that thegas first encounters the blade while flowingaxially. This allows increased head output andhigher stage pressure ratios. The use of three orfour stages is more typical nowadays. Theconfiguration might consist of two LP stages ona rotor running at 3,000 to 5,000 rpm, and twoHP stages on a smaller, shorter rotor running at4,000 to 11,000 rpm.

Mechanical gas extractors were presumablychosen at Wairakei because the extra steamrequired to generate the electricity to drive themis approximately one quarter of the steamneeded to drive the equivalent capacity steam jetejectors. However, this figure exaggerates thebenefit of mechanical gas extractors, as thevalue of the steam to drive the ejectors is muchless than the displaced electricity generationinherent in a mechanical gas exhauster. This isbecause geothermal power generationeconomics is largely capital cost driven, and the

steam generation (steamfield) costs are typicallyless than half of the capital cost of the entireproject.

However, at high gas contents the steamconsumption of steam jet ejectors gets to be solarge that mechanical gas exhausters becomemuch more common. For example, Ohaaki,which has a gas content of approximately 4%,selected mechanical gas exhausters. Theseconsume about 2 MW of a 40 MW gross turbineoutput, compared to steam jet ejectors whichwould have consumed about 8 MW equivalentof steam.

Although steam jet ejectors and mechanical gasextractors are both very inflexible with respectto gas flow, steam jet ejectors gain flexibilitythrough their relatively low capital cost. Thisallows installation of 2x100% or 3x50% systemsto give 50%, 100% and 150% capacity forexample. Ejectors are simply turned on or off asrequired. The high capital cost of mechanicalgas extractors tends to promote 1x100% or2x50% designs. Less than 100% flows can beaccommodated using recirculation, but at greatlyreduced efficiencies – typically very little drivepower reduction is achieved even with only 25%of design flow. Although it has not been used ingeothermal applications to date, variable speeddrives are used in other industries oncompressors to give efficient turndown to thesurge limit (about 50% of design flow). Newermodular compressor designs allow flexibility fora particular casing configuration at the selectionstage, but post-installation capacity changeswould still require impeller replacement as wellas piping reconfiguration. As the moderninducer impeller requires a 3 dimensionaldesign, they form a significant part of theexhauster cost, and therefore re-rating isexpensive.

Outcomes from typical project economicevaluations for a modern application produce thefollowing rules of thumb. Steam jet ejectors(typically 3-stage for maximum efficiency) areusually preferred for a non-condensable gas(NCG) content up to 1.5%. Hybrid systemsconsisting of steam jet ejectors for the initialstages and liquid ring vacuum pumps (positivedisplacement) for the final stage are preferredfor a NCG content between 1 and 3%.Mechanical gas exhausters are preferred for aNCG content greater than 3%. On this basis,steam jet ejectors would probably be selected fora new application at Wairakei because of thelow gas content. Where gas contents are higher,mechanical gas extractors based on successful

Italian experience, continue to be used withrelatively little trouble.

6. CONCLUSIONS AND RECOM-MENDATIONS

Wherever possible, the original designers of theWairakei Power Project tried to employestablished technology with some specificrefinements. A number of the technologiesrejected prior to or shortly after commissioningof the Wairakei Power Station in 1958 havebeen looked at again for this paper. Some, suchas the heavy water plant have limited potential.Others, such as centralised multi-flash systems,binary cycle plant and mechanical gas extractionsystems have achieved common usage inapplicable situations. Simple technology such asthe U-bend separator has been overlooked buthas the potential for capital savings particularlyin relation to binary cycle or process heatapplications, with a little more research.

While a reasonably conservative approach todesign is needed to aid funding of geothermalprojects, there is an obligation on designers tokeep an open mind on potential technologies.Old ideas should be reviewed to see if theycould work in today’s context. Other industriesshould be watched to allow cross-fertilisation ofideas. Wairakei Power Project was a soundlybased project, but future developments will havea different blueprint.

ACKNOWLEDGEMENTS

The authors would like to thank the followingorganisations: PB Merz and McLellan (NZ) Ltd(formerly DesignPower New Zealand Ltd),ECNZ Ltd (for library assistance) and ContactEnergy Ltd (for review and permission topublish Wairakei information). We would like tothank the following individuals for personaladvice and research assistance: Dick Bolton,Tom Lumb, Bill McCabe, Tony Mahon andHilel Legmann.

REFERENCES

AJHR (1956). Annual Report, Engineer-in-Chief, MOW. In: Appendix to the Journal of theHouse of Representatives, 1956, Section D1,pp45-46.

Angelino, G., Gaia, M. and Macchi, E. (1984).A Review of Italian Activity in the Field of

Organic Rankine Cycles. Proceedings of theInternational VDI – Seminar held in Zurich 10-12 September, 1984, 465-482.

Aronson, D. (1961). Binary Cycle for PowerGeneration. Proc. Of Amer. PowerConference,23 1961.

Bolton, R.S. (1977). Geothermal Energy in NewZealand. In: Energy Technology Handbook,Considine (Ed.), McGraw-Hill, pp7-14 to 7-31

Bolton, R.S. (1998). Notes on the Early Historyof Wairakei. Proceedings 20th New ZealandGeothermal Workshop 1998.

Bronicki, L.Y. (1984). Twenty Five YearsExperience With Organic Working Fluids inTurbomachinery. Proceedings of theInternational VDI-Seminar held in Zurich 10-12September, 1984, 685-696.

Crossland, I.G., Myerscough, P.B., Palmer, J.E.,Pugh, A.R., Simpson, K.A. and Streatfield, R.E.(1992). Nuclear Physics and Basic Technology.In: Modern Power Station Practice, Volume J,Nuclear Power Generation, P. B. Myerscough(Ed)., Third Edition, Pergamon Press.

DiPippo, R (1980). Geothermal Power PlantsAround the World. In: Sourcebook on theProduction of Electricity from GeothermalEnergy, J. Kestin (Ed)., USDOE.

Glasstone, S. (1967). Sourcebook on AtomicEnergy. Third Edition, Van Nostrand, Princeton.

Haldane, T.G.N. and Armstead, H.C.H. (1962).The Geothermal Power Development atWairakei, New Zealand. Proc Instn Mech Engrs,Vol 176 No 23 1962, 603-649.

Love, R.N and Bolton, R. (1980) WairakeiGeothermal Power Project: New Zealand. In:Geothermal Energy Projects Planning andManagement, L.J. Goodman and R.N. Love(Ed)., Pergamon Press.

Luchter, S. (1967). Current Status of theTechnology of “Organic” Rankine Cycle PowerPlants. Proc. Intersociety Energy ConversionEngineering Conference- “Advances in energyconversion engineering” – Miami, FL, 13-17August 1967, 117-121

Smith, J.H. and McKenzie, G.R. (1970).Wairakei Power Station New Zealand –Economic Factors of Development andOperation. Geothermics (1970) – Special Issue,1717-1723.

Smith, J.H. (1973). Collection and Transmissionof Geothermal Fluids. Unesco, 1973.Geothermal Energy (Earth Sciences, 12.), 97-106.

Stacey, R.E. and I.A. Thain (1983). Twenty-FiveYears of Operation at Wairakei GeothermalPower Station. Unpublished Draft Report forElectricity Division – Ministry of Energy.

Tabor, H. (1961). Large Area Solar Collectors(Solar Ponds) For Power Production. UNConference, New Sources of Energy Rome 1961.

Zhou, Hong-Yi (1983). Binary Cycles in China.Proc 5th NZ Geothermal Workshop 1983, 275-279