Energy optimisation of existing SWRO (seawater reverse ... · Energy optimisation of existing SWRO (seawater reverse osmosis) ... (energy recovery turbines): Technical and thermoeconomic

lable at ScienceDirect

Energy 36 (2011) 613e626

Contents lists avai

Energy

journal homepage: www.elsevier .com/locate/energy

Energy optimisation of existing SWRO (seawater reverse osmosis) plants with ERT(energy recovery turbines): Technical and thermoeconomic assessment

Baltasar Peñate a,*, Lourdes García-Rodríguez b,1

aWater Department, Canary Islands Institute of Technology (ITC), Playa de Pozo Izquierdo s/n, 35119 Santa Lucía - Las Palmas, SpainbDepartamento de Ingeniería Energética, Universidad de Sevilla, ETSI, Camino de Los Descubrimientos s/n, 41092-Sevilla, Spain

a r t i c l e i n f o

Article history:Received 5 May 2010Received in revised form27 September 2010Accepted 30 September 2010Available online 30 October 2010

Keywords:Pelton turbines retrofitReverse osmosis seawater desalinationLowest energy consumptionHigher capacityThermoeconomic assessment

* Corresponding author. Tel.: þ34 928727511; fax:E-mail addresses: [email protected] (B.

(L. García-Rodríguez).1 Tel.: þ34 954487231; fax: þ34 954487233.

0360-5442/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.energy.2010.09.056

a b s t r a c t

Reduction of SEC (specific energy consumption) is the field with the most specific technical researchfocus and effort in SWRO (seawater reverse osmosis) plants. For existing installations with energyrecovery systems consisting in Pelton turbines, the most significant challenge is how to reduce energycosts. The highest efficient isobaric ERD (energy recovery devices) are used in order to produce majorsavings in energy consumption in the desalination process and/or to increase the freshwater capacity ofthe installations, by taking full advantage of the plant equipment. This paper gives a brief overview of thetechnology used to recover the energy from brine stream in large desalination plants, with a descriptionof the modifications required if the recovery system with Pelton turbines is to be replaced by systemsbased on isobaric-chamber devices. All possibilities analysed are deeply justified technically and ther-moeconomically within an exhaustive assessment.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

RO (Reverse Osmosis) desalination plants have become themajor focus for engineering and technological innovation overthe last few decades, representing one of the areas of greatestgrowth. SWRO (Seawater Reverse Osmosis) technology is now allover the world with the major capacity installated respected therest of desalination technologies [1]. The SEC (Specific EnergyConsumption) reduction has monopolised the focus of technolog-ical innovation and research in this sector. The energy costs inSWRO plants could represent up to 50% of the final costs of thewater product, thus making highly efficient Energy RecoveryDevices (ERD) of vital importance, in that they allow for energy tobe recovered from the brine stream.

The first design carried out in the early Eighties used systemsbased on a centrifugal pump, an engine and either, Francis or Peltonhydraulic turbines. These systems allowed for SECs less than5.00 kWh/m3 [2]. At present, large desalination plants are usingisobaric-chamber devices which are over 95% energy transfer effi-cient and offer SECs below 2.50 kWh/m3 [3,4]. Obviously, there has

þ34 928727590.Peñate), [email protected]

All rights reserved.

been more and more research ploughed into new and bettersystems. Nevertheless, the classic system of energy recovery fromthe brine stream via turbines, in spite of it being improved andoptimised over the years, is not expected to offer greater energytransfer efficiency. The present trend is toward replacement byisobaric-chamber devices to achieve energy savings in large plants.

Managers of existing SWRO plants are thus faced with thedelicate and difficult decision to opt for retrofitting of the ERT(Energy Recovery Turbines). It permits to augment the useful life-span of the plant and achieve greatest reductions in energyconsumption over the desalination process [3,5]. Other possibilitycould be trying to increase the output capacity of the plant bytaking maximum advantage of the equipment in place [6]. In theCanary Archipelago alone, which is all of a European reference inthe field of desalination plants with one of the highest density inthe world of SWROs (22.73 km2/plant) [7,8], there are over 130seawater desalination plants with a total installed capacity of over400,000 m3/d [9]. Most of the large public owned plants wereset up before isobaric-chamber energy savings devices existed,which is the reason for which they have the defect ERT systemsinstalled. Therefore, the future of these plants lies in their graduallyoptimising and updating their energy recovery systems [7].

The changes to be carried out in the default type system in caseof retrofitting are described in the following sections. This retrofitassessment includes technical and thermoeconomic analysis of theoptions proposed.

Table 1Energy recovery turbines vs. isobaric energy recovery devices: a generalcomparison.

Criteria Pelton turbines Isobaric ERDs

Typical train capacity Up to 5000 m3/d More than 5000 m3/dTypical recovery rate 45e50% 37e45%HPP power For pumping

100% feed waterFor pumping 37e45%of total feed water

No. of booster pumpsrequired

n/a 1 for each energyrecovery device

Relevant civil workrequirements

Yes No

Salinity increase of thefeed water beforemembranes

No Yes

Extra brine pipeinstallation

Yes No

Low pressure differentialin the energy recoveryprocess

n/a Yes

High pressure differentialin the energy recoveryprocess

n/a Yes

Flow leakages (lubrication) n/a YesAverage efficiency Up to 90% 90e97%Typical SEC achieved 3e4 kWh/m3 2e3 kWh/m3

Energy savings range(after installing)

35e42% 55e60%

n/a: not applicable.

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626614

2. ERD for SWRO plants

Several devices have been designed and tested to recover theenergy from the brine flow within SWRO. The first ERD used wasa turbine coupled to the centrifugal HPP (High-Pressure Pump)shaft. Before the 1980s, Francis turbines were applied, but theywere replaced by the Pelton wheel. The latter technology operatesat higher efficiency in high-head applications such as large SWROplants. ERTs have been widely accepted in large SWRO plants

Fig. 1. 10,000 m3/d SWRO plant diagram usin

because of their reliability and proven efficiency, typically up to 88%[10].

The whole component efficiencies of a Pelton turbine ERD rangefrom 70% to 90%. The energy transfer efficiency of this kind of ERDis the product of the efficiencies of the nozzles (99%), the turbine(85%e90%) and the centrifugal HPP (75%e85%) [11]. In conclusion,the peak efficiency of a Pelton turbine ERD could be estimatedaround 85e90%.

The operation data monitoring of an ERT, e.g. flows, pressures,RPM and efficiency of each devices, is inter-related in a real situa-tion. So that, it produces a host of conflicts and problems forreaching an optimum operation point [12]. Besides, at lowerrecovery rates, the ERTs are pumping more feed water. The onlyway to reduce the energy costs is to increase the recovery rate ofthe RO plant. Therefore, the solution for current installations withERTs is now being designed at membrane-challenging recoveryrates higher than 45%.

Another type of centrifugal ERD is the hydraulic turbocharger,which was used for small and medium capacity SWRO plants atthe beginning of the 1990s. They are similar in operational conceptto the Pelton turbine. The turbocharger and the HPP are not directlyconnected, providing a degree of flexibility in the operation of thesedevices. Also, turbochargers have a relatively small footprint andare easy to install but the overall efficiency of this ERD is typically70e80%.

From the 1990s onward, several alternatives to centrifugal ERDhave been designed and tested. These kinds of devices use theprinciples of isobaric chambers for SWRO plants and are known asisobaric pressure exchangers [13].

The HPP size and its motor is substantially smaller than ERTconfiguration. It should be more exactly to say that this kind ofdevice saves energy instead of recovering energy as ERT.

There are several manufacturers of pressure exchanger devicesnowadays. ERI� (PX) [13], Calder (DWEER�) [14], RO Kinetic� [15]and KSB (SalTec DT) [16], Danfoss (iSave) [17] are the main manu-facturers of this kind of equipment and are now competing for

g Pelton turbine energy recovery devices.

Table 2Table 22 � 5000 m3/d SWRO plant train characteristics.

Description per train Train i (i ¼ 1,2)

Seawater feed 463 m3/hPermeate stream 208 m3/hBrine water stream/turbine inlet 255 m3/hNo. vessels 52No. elements 364Feed flow per vessel 8.9 m3/hBrine flow per vessel 4.9 m3/hPermeate flow per vessel 4.0 m3/h

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626 615

designing the best system. Some optimised devices allow a realefficiency of energy from the brine stream to the feed flow of up to97%. Besides, the innovative ERDs for SWRO plants imply an enor-mous advance in the SEC reduction in this type of facilities. Thesedevices obtain excellent SEC from 1.80 to 2.20 kWh/m3 in newmedium capacity SWRO plants with piston HPP installed [4,15].

A general comparison between standard ERT and isobaricERDs trains for large capacity SWRO plants is shown in the Table 1.The data does not try to compare existing plants but the maindifferences at the moment of designing both.

3. Pelton turbine retrofit

As previously stated in the preceding section, all the currentERDs based on isobaric chambers reduce the RO energy consump-tion in comparison to ERT. For instance, in a case of an SWRO plantwith ERT and a SEC of 4.00 kWh/m3, where a new generation ERDcould be installed with a total SEC of 2.35 kWh/m3, the energysavings is close to 41%. Obviously, the reduction of the SEC repre-sents an excellent advantage for our installation [3]. However, ina preliminary selection of a new isobaric ERD to replace an ERTdevice, other relevant aspects should be taken into consideration:

� Overall efficiency.� New operation data and recovery rate.� Civil works, hydraulic pipes and electrical modifications.� Capital investment required.� Operation andmaintenance requirements and worker training.

Therefore, a general comparison between isobaric ERDs has tobe carried out in order to narrow down the final selection. Likewise,and parallel to this comparison, it is needed to analyse the retrofitconfigurations of the RO membrane racks to target a theoreticalincrease in the recovery rate, whilst maintaining the HPP andmotoror modifying these to adapt to the new flows. There are variouspossibilities available. The most suited and feasible configurationsare described in the following sections.

For the design of each configuration, a calculation has beenmade using well-intake seawater of 35,000 TDS at 20 �C and pH7.50. It has taken the HPP inlet as 2.5 bar with seven-element PV(pressure vessel) with SW30XLE 400i membranes, 0.85 fouling

Table 310,000 m3/d SWRO plant data operation with two standard Pelton turbinesconfiguration.

Description Flow (m3/h) Pressure (bar) Pipe dimension

Total feed 926 n/a n/aHPP inlet 463 2.5 300 mmRO train feed 463 58.6 203 mm (800)Brine water stream/turbine inlet 255 57.0 203 mm (800)Turbine outlet/brine discharge 255 0.00 300 mmPermeate stream per train 208 n/a n/a

n/a: not applicable.

factor, 0.78 HPP performance, 0.84 turbine efficiency, 0.77 BOPperformance, 0.98 Variable Frequency Drive (VFD) performanceand 0.92 motor efficiency.

An isobaric ERD was designed according to the followingcriteria: 0.97 ERD efficiency, 0.7 bar ERD LP differential, 1.1 barERD HP differential and 1.3% lubrication flow. Piping and valvelosses were considered and permeate back pressure was ignored.Filmtec RO design software Rosa v.7.01 [18] was used to simulatethe PV.

3.1. Standard configuration with Pelton turbine energy recovery

Fig.1 shows a 2� 5,000m3/d SWROplantwith two ERT installed.Feed water from the intake is fed over the existing identical trains.The operational pressure is attained thanks to the HPP which iscoupled to a Pelton turbine on the same shaft. The energy generatedby the Pelton turbine allows the energy consumption of the HPP tobe reduced considerably.

In the case of this simulation, each train produces 5000 m3/d at45% recovery and 58.6 bar feed pressure. The pressure loss in thePVs and pipes is 1.6 bar, so the pressure of the brine stream in theturbine inlet is 57.0 bar.

Tables 2e4 show the whole operational data of this configura-tion, based on the design criteria described above.

The energy requirements of each turbopump stand at around681.8 kW, so that, the total energy requirement of the installation is1363.6 kW with a SEC of 3.27 kWh/m3.

Real experiences of SWRO plants with this kind of ERTs showthat the energy consumption may be substantially higher. It is thecase of the 80,000 m3/d Las Palmas III SWRO plant (The CanaryIslands - Spain). Some of its 9 trains operated with ERT (Peltonturbines) from 1998 through to the beginning of 2008 and the SECranged between 3.35 and 3.50 kWh/m3 [19]. This installation wasretrofitted with six-element PVs using Hydranautics SWC1 andSWC3þ membranes.

The current situation of the 30,000 m3/d Lanzarote IV SWROplant (Spain) is similar: 6 trains with seven-element PVs usingFilmtec SW30-380 elements - with an ERT (Pelton turbine) andSEC of 3.42 kWh/m3 [20] from 2000.

4. Constant capacity configuration to reduce energyconsumption

4.1. Retrofit No.1: the same HPP and installation of an isobaricenergy recovery device

This configuration is recommended for cutting the energyconsumption with the same total plant capacity. The Peltonturbines are discarded and one of the existing HPPs is used forevery two trains. This configuration requires replacement of theHPP motor should this not cover the total energy demanded,together with certain modifications, in some cases, of sections ofelectric wiring and medium-voltage electric protections. Thesemodifications do not tend to be a problem in the MVTS (Medium-Voltage Transformer Substation) since half of the existing HPPs are

Table 4Energy requirements in a 10,000 m3/d SWRO plant with two standard Peltonturbines configuration.

Description Train i (i ¼ 1,2) Total

HPP (kW) 966.6 1933.3Pelton Turbine (kW) (�) 339.4 (�) 678.8Motor energy requirement (kW) 681.8 1363.6SEC 3.27 kWh/m3

Fig. 2. Retrofit proposed in an existing SWRO plant - Isobaric energy recovery device in a two 5.000 m3/d RO trains.

Table 610,000 m3/d SWRO plant comparative pipe dimensions (standard Pelton turbineconfiguration (ERT) vs. isobaric energy recovery device (ERD) configuration).

Description Pipe dimensionERT � 1 train

Pipe dimensionERD

Seawater feed (A) n/a n/aHPP inlet (C) 300 mm 300 mmERD feed (B) n/a 300 mmBOP feed (D) n/a 254 mm (1000)RO train feed (F) 203 mm (800) 356 mm (1400)Brine water stream/Turbine

inlet/ERD HP feed (G)203 mm (800) 254 mm (1000)

Brine discharge Turbine & ERDLP outlet (H)

300 mm 300 mm

n/a: not applicable.

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626616

in use and thus the global consumption at medium-voltage willbe lower than the option with turbopumps. Therefore, the MVTSshould not require extensions.

The configuration does not involve changing or extending thelow-pressure seawater feed pumped. It demands the same amountof feed water per train. Neither the recovery rate nor the flowsare modified. However, this retrofit requires a modification ofhydraulic installation due to the fact that one ERD is installed everytwo trains, whichwill require BOP and newchannelling of the flowsof feed water and brine.

By introducing BOPs, whose motors work at low-voltage withVFDs, there will be a need to increase the LVTS (Low-VoltageTransformer Substation). Fig. 2 gives a basic layout of this config-uration, using one of the existing ERTs. The configuration requiresbarely additional space since the ERDs can be installed on the sametrain in the area left by the Pelton turbines.

Tables 5e7 show the whole operational data of this configura-tion, based on the design criteria described above.

Table 5Comparative operational data for a 10,000 m3/d SWRO plant (standard Peltonturbine configuration (ERT) vs. isobaric energy recovery device (ERD) configuration).

Description Flow ERT � 1train (m3/h)

Flow ERD(m3/h)

PressureERT � 1train (bar)

PressureERD (bar)

Seawater feed (A) 463 926 n/a n/aHPP inlet (C) 463 416 2.5 2.5ERD feed (B) n/a 510 n/a 2.5BOP feed (D) n/a 510 n/a 55.5RO train feed (F) 463 463 58.6 58.6Permeate 1 n/a 208 n/a n/aPermeate 2 n/a 208 n/a n/aPermeate stream (I) 208 416 n/a n/aBrine water stream/Turbine

inlet/ERD HP feed (G)255 510 56.6 56.5

Brine discharge Turbine &ERD LP outlet (H)

255 510 0 1.2

n/a: not applicable.

The recovery rate is constant over both trains, but the flows arenow distributed over the plant in different ways from the original.The feed flow is divided into streams A and B. The high pressurevalue is the same over the two trains. An additional requirementto be borne inmind is that the HPP (stream C) now impels less flow.It should be necessary for a pump retrofit, to eliminate some of thestages.

Table 7Energy requirements in a 10,000 m3/d SWRO plant (standard Pelton turbineconfiguration (ERT) configuration vs. isobaric energy recovery device (ERD)configuration).

Description ERT configuration ERD configurationRetrofit No. 1

HPP (kW) 1933.3 946.5Pelton Turbine (kW) (�) 678.8 n/aBOP(kW) n/a 67.7Total requirement (kW) 1363.6 1014.2SEC (kWh/m3) 3.27 2.43Energy savings (%) 25.6

n/a: not applicable.

Fig. 3. Retrofit proposed in an existing SWRO plant - Isobaric energy recovery device in a two 5.000 m3/d RO trains and a booster pump for the low-pressure feed water.

Table 9Energy requirements of a 10,000 m3/d SWRO plant (standard Pelton turbineconfiguration (ERT) vs. isobaric energy recovery device (ERD) configuration witha booster pump (BOP) in the feed water stream).

Description ERT configuration ERD configurationRetrofit No. 2

LP feed BOP (kW) n/a 279.7HPP (kW) 1933.3 681.8Pelton Turbine (kW) (�) 678.8 n/aBOP (kW) n/a 67.7Total requirement (kW) 1363.6 1029.2SEC (kWh/m3) 3.27 2.47Energy savings (%) 24.5

n/a: not applicable.

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626 617

Some pipe diameters have to be modified and new pipes areneeded. The new installation needs an additional pipe for the BOPstream.

From the above table, it can be clearly seen that there is a needtomodify the motor of the HPP. The energy required increases from681.8 kW to 946.5 kW. The total energy requirement in thisconfiguration is 1014.2 kW, so energy savings of 25.6% is obtainedwith a reduction of the SEC by 0.84 kWh/m3 in this application.This energy saving allows for a quick payback period of the newequipments installed.

An excellent example of this type of retrofit has been imple-mented in the Mazarrón (Spain) SWRO plant. This SWRO plantcombined two 3000 m3/d trains in 2001 and installed ERI brandERDs as replacements of two Pelton turbines. A reduction of0.73 kWh/m3 was obtained plus savings of 27%. The SEC wasreduced to 2.37 kWh/m3 [21].

Another experience is the retrofit made in the 40,000 m3/d Dhekelia SWRO plant (Cyprus). A retrofit combined 2 � 5000 m3/d trains into a single 10,000 m3/d train. In this case, six Francisturbines were replaced by three rack ERDs (ERI technology) indifferent phases from 2001 to 2004 [22]. In late 2008, Las PalmasIII (The Canary Islands-Spain) combined two trains into a single15,000 m3/d train with ERI ERDs. The SEC was reduced from3.49 kWh/m3 to 2.68 kWh/m3 [13].

A minor modification of this configuration would consist inslightly increasing the train capacity installed, from 5000 m3/d to5556 m3/d. Thus, the water feed flow passing through the HPP(stream C) would coincide exactly with the feed flow through theoption using ERT (463 m3/h instead of 416 m3/h). Thus, dependingupon the specifications of the HPP, a retrofit which involves

Table 8The comparative operational data for a 10,000 m3/d SWRO plant: standard Pelton turbinea booster pump (BOP) in the feed water stream.

Description Flow ERT � 1 train (m3/h) Flow ERD þ low pressure BOP (m

Seawater feed (A) 926 926LP BOP inlet (A) n/a 416HPP inlet (C) 463 416ERD feed (B) n/a 510RO train feed (F) 463 463

n/a: not applicable.

eliminating some hydraulic stages can be avoided. This modifica-tion allows for increasing the capacity (11,112 m3/d), which wouldslightly improve the SEC of the installation which would drop to2.23 kWh/m3, but consuming slightly more energy. There wouldbe a total energy saving of 24.1%. It should be noted that thisretrofit could be an awkward option to implement since morePVs are required and there is no space, apart from the fact thatenergy demands are higher at medium-voltage.

4.2. Retrofit No.2: isobaric energy recovery device with a lowpressure BOP to increase the low feed water pressure

By way of modification of the previously described case and toavoid replacement of the existing motors already coupled to theHPP, a low pressure BOP is installed previous to the HPP. In Fig. 3,

(ERT) configuration vs. an isobaric energy recovery device (ERD) configuration with

3/h) Pressure ERT � 1 train (bar) Pressure ERD þ low pressure BOP (bar)

n/a n/an/a 2.52.5 18.5

n/a 2.558.6 58.6

Fig. 4. Installation of an isobaric energy recovery device and higher RO train.

Table 10Comparison between 5000 m3/d and 7200 m3/d RO trains.

Description per train 5000 m3/d RO train 7200 m3/d RO train

Seawater feed 463 m3/h 667 m3/hPermeate stream 208 m3/h 300 m3/hBrine water stream 255 m3/h 367 m3/hRecovery rate 45% 45%No. vessels 52 75No. elements 364 525Feed flow per vessel 8.9 m3/h 8.9 m3/hBrine flow per vessel 4.9 m3/h 4.9 m3/hPermeate flow per vessel 4.0 m3/h 4.0 m3/h

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626618

the layout of this combination is given. As can be seen, there is nochange in the part of the installation relating to the isobaric ERDand the high pressure BOP in comparisonwith the previous retrofitdescribed.

The newly fitted BOP will be responsible for raising the pressureof the feed water to the value which is required, energetically, tomaintain a constant flow to the existing HPP motor. This avoidssubstantial electrical modifications at medium-voltage due to theneed to increase the motor power, together with the change in theelectric wiring and existing electric protections. No further space isrequired for new trains since the total installed capacity remainsthe same.

Table 11Comparative operational data for an SWRO plant after the installation of an isobaricenergy recovery device (ERD) to increase the train capacity.

Description Flow ERT(m3/h)

Flow ERD þ higherRO train (m3/h)

PressureERT (bar)

PressureERD þ higherRO train (bar)

Seawater feed (A) 463 667 n/a n/aHPP inlet (C) 463 300 2.5 2.5ERD feed (B) n/a 367 n/a 2.5Booster pump

feed (D)n/a 367 n/a 56.9

RO train feed (F) 463 667 58.6 58.7Permeate stream

(I)208 300 n/a n/a

Brine waterstream/Turbineinlet/ERD HPfeed (G)

255 367 56.6 57.2

Brine dischargeTurbine &ERD LP outlet (H)

255 367 0 1.2

n/a: not applicable.

Likewise, since new BOPs are introduced, the motors of whichwork at low-voltage with VFD, there will be a need to increase theLVTS and MVTS.

It is recommended that a hydraulic retrofit be made to theHPP at the inlet part since it should pump feed water at mediumpressure. In Tables 8 and 9 the operating values according to thepractical case of 10,000m3/dwhich is being carried out can be seen.

This simulation allows deducing that a BOP in series with HPPshould be capable of raising the feed water flow from 2.50 to18.50 bar. The HPP retrofit produces a pressure increase from18.5 bar to 58.6 bar.

This design offers a 19.5% increase in the low-voltage energyrequired as compared to the previous retrofit. There are worseSEC and less energy savings but, nevertheless, the results are betterin comparison with the initial installation with ERT, producinga reduction of around 0.8 kWh/m3. It should be borne in mind thatthis configuration is designed to minimise changes and, thus, insome cases, may be advisable.

5. Configurations to increase the capacity

As opposed to the previous retrofits, where the main criterionwas to bring down global energy consumption over the installationwithout considering increased production as a valid design crite-rion, in this section are described configurations focused onincreasing the installed capacity. As a consequence, the plantreduces its SEC although more energy is consumed but the ratio ofenergy invested to obtain water is reduced. These retrofits arespecifically recommended in installations with ERT where thereis space to increase the number of PVs per train and there isinsufficient capital to allow for the investment in new installations.

Table 12SWRO plant comparative pipe dimensions after the installation of an isobaric energyrecovery device (ERD) to increase the train capacity.

Description Pipe dimension ERT Pipe dimensionERD þ higher RO train

Seawater feed (A) n/a n/aHPP inlet (C) 300 mm 300 mmERD feed (B) n/a 300 mmBOP feed (D) n/a 203 mm (800)RO train feed (F) 203 mm (800) 254 mm (1000)Brine water stream/Turbine

inlet/ERD HP feed (G)203 mm (800) 203 mm (800)

Brine discharge Turbine &ERD LP outlet (H)

300 mm 300 mm

n/a: not applicable.

Table 13Energy requirements in an SWRO plant after the installation of an isobaric energyrecovery device (ERD) to increase the train capacity.

Description ERT configuration ERD þ higher RO trainconfiguration Retrofit No. 3

HPP (kW) 966.6 662.3Pelton Turbine (kW) (�) 339.4 n/aBOP (kW) n/a 54.7Total requirement (kW) 681.8 717.04SEC (kWh/m3) 3.27 2.39Production increase (%) 44.2

n/a: not applicable.

Table 147693 m3/d capacity SWRO plant train characteristics.

Description per train Train 1 (5.000 m3/d) Train 2 (2.693 m3/d)

Seawater feed 463 m3/h 255 m3/hPermeate stream 208 m3/h 112.2 m3/hBrine water stream 255 m3/h 142.8 m3/hRecovery rate 45% 44%No. vessels 52 31No. elements 364 217Feed flow per vessel 8.9 m3/h 8.2 m3/hBrine flow per vessel 4.9 m3/h 4.6 m3/hPermeate flow per vessel 4.0 m3/h 3.6 m3/h

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626 619

5.1. Retrofit No.3: higher capacity per train and installation of anisobaric ERD

This configuration implies taking advantage of the existingtrains and increasing the number of PVs to achieve an increase inthe total installed capacity. HPP continues to be used together withthe same motor but the Pelton turbine is disconnected from theshared shaft.

The parameters used for defining the new capacity of existingtrains are the power of the existing motor and the value of the highpressure operation of the existing plant with ERT. Fig. 4 shows thelayout of this retrofit proposed.

In this design, the HPP is retrofitted to allow for 35% feed flowreduction. The design is completed with an isobaric ERD anda BOP to increase the feed water pressure at the ERD outlet. Thisconfiguration consumes more net energy but allows for 40% morepermeate water.

This design requires modification or increasing the feed waterintake pumps together with new seawater beach wells if necessary,since more seawater is required per train. The pipes of the plantalso suffer slight modifications since a new larger diameter feedlineto the train (F) is required or parallel feeding is provided via a newpipe to the new PVs of each train. Likewise, as more water is beingproduced, the section of permeate piping must be increased ormore parallel pipes be installed.

Fig. 5. Isobaric energy recovery device

As in the previous cases, once the BOPs are installed and theneed of seawater feed increases, the pumps whose motors work atlow-voltage with VFDs will need necessary to increase the LVTS.

One drawback of this configuration is that it requires space toincrease the number of PVs and to install ERDs together with BOPs.

In the practice, it has seen that the former 5000 m3/d train usedin the ERT configuration increases to a new 7.200 m3/d train. Tables10e13 show thewhole operational data of this configuration, basedon the design criteria described above.

The new train requires 23 new PVs and works at identicalrecovery rate. The energy necessary to raise the 300 m3/h pumpedby the HPP to 58.7 bar is equivalent to the energy which iscontributed by the existing motor connected to the HPP.

From the perspective of piping, it is only required to increasethe diameter of the piping related to the RO train inlet pipes. It isalso necessary to position a new hydraulic pipe for the low andhigh-pressure feed flow to the ERD and the BOP.

In this configuration, the same net energy is consumed with anincreasing of 44.2% in water production. As a result, the SEC ofthe installation can be reduced by some 0.88 kWh/m3. It shouldbe borne in mind that the energy required for the feed water intakeis higher than in previous cases so that the overall SEC is slightlyincreased due to this new energy requirement.

An example of this retrofit is proposed in the Via Maris Palma-chim SWRO plant (Israel). The capacity of the plant is 120,000 m3/

with a new RO train without BOP.

Table 15Comparative operational data for an SWRO plant after installation an isobaric energy recovery device (ERD) used as a Booster pump to increase the train capacity.

Description Flow ERT (m3/h) Flow ERD e new train (m3/h) Pressure ERT (bar) Pressure ERD e new train (bar)

Seawater feed 463 718 n/a n/aHPP inlet (C) 463 463 2.5 2.5ERD feed (B) n/a 255 n/a 2.5BOP feed (D) n/a n/a n/a n/aRO train 1 feed (C) 463 463 58.6 58.6RO train 2 feed (D) n/a 255 n/a 56.1Permeate stream (H) 208 208 n/a n/aPermeate stream (I) n/a 112 n/a n/aBrine water stream/Turbine inlet/ERD HP feed (E) 255 255 56.6 57.2Brine discharge Turbine & ERD LP outlet (F) 255 255 0.00 1.2Brine discharge train 2 (G) n/a 143 n/a 0

n/a: not applicable.

Table 16Energy requirements in an SWRO plant after isobaric energy recovery device (ERD)used as a Booster pump to increase the train capacity.

Description ERT configurationper train

ERD RetrofitNo.3

ERD e new trainRetrofit No.4

HPP (kW) 966.6 662.3 966.6Pelton Turbine (kW) (�) 339.4 n/a n/aBOP (kW) n/a 54.7 n/aTotal requirement (kW) 681.8 717.04 966.6SEC (kWh/m3) 3.27 2.39 3.02Total capacity (m3/d) 5.000 7.200 7.693Production increase (%) n/a 44.2 53.8%

n/a: not applicable.

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626620

day in 6 � 20,000 m3/day RO trains. After the retrofit a reduction of18% in the SEC is expected [23].

5.2. Retrofit No.4: higher capacity per train and installation of anisobaric ERD used as a BOP

A modified version of the previous case consists in using theisobaric ERD as if it were 2nd-stage HPP. MacHarg (2002) [12] callsthis retrofit cascade expansion. In Fig. 5, it can be seen howa portion of the feed water goes into the isobaric ERD and is madeto acquire the pressure of the 1st-stage brine rejected. The high-pressure feed flow is passed over a new train designed to that

Fig. 6. Isobaric energy recovery device with a new RO

effect. In general, this alternative produces more water than theprevious configuration - more than 50% and allows for the existingHPP and RO train to be used. No further installation of energyconsuming equipment is required (BOP). However, this configu-ration is not as efficient, energy-wise, as the previous configura-tion and requires that the ERT motor be changed for a highermotor power, besides needing more space since a new RO trainmust be installed.

Tables 14e16 show the whole operational data of this configu-ration, based on the design criteria described above.

As in the original configuration, 100% of the feed water flowsthrough the HPP which makes it necessary to change the motor,for a total recovery rate over the whole process of 44.5%. Thesecond-stage feed pressure is 57.2 bar.

In comparison with previous retrofit, despite increasing thecapacity this retrofit increases the SEC of the installation because itproduces an increase in the net energy consumption required fordesalination. As compared to an ERT configuration, there is a slightreduction in the SEC (0.25 kWh/m3) although what is outstandingis the 53.8% increase in the water product.

Few real experiences are known with this configuration. The St.Croix SWRO plant (US Virgin Islands) is an example of cascadeexpansion. 83 m3/d at 64.8 bar feed pressure and 39% recovery ratewas modified with a new train of 34 m3/d at 63.8 bar feed pressure.A 41% production increase was obtained at 4.1 kWh/m3 [12].

train without booster pump and Pelton turbine.

Table 17Exergy flows (Fuel, Product and Losses) of the equipments.

Equipments Exergy Flows

Fuel (F) Product (P) Losses (L)

Equipment 0 E�

x0 þ E�

x4ð¼ PWauxÞ E�

x1 e

Equipment 1 E�

x1 þ E�

x5ðPWbÞ E�

x3 ¼ E�

xP E�

x2 ¼ E�

xbd

Fig. 7. Flow chart of the whole productive process for the analysis in the case ofstandard configuration (existing desalination plants with energy recovery device basedon turbine Pelton).

Table 18Exergetic costs and unitary exergetic costs of the flows.

Flows (j) Exergetic costs

E*xj (kW) cxj

0 Ex*0 ¼ qf $exf ¼ 0 cx0 ¼ 0

1 Ex*1 ¼ Ex*0 þ PWaux ¼ PWaux cx1 ¼ PWaux

qf $exðbf ;Tf ;pf1Þ2 Ex*2 ¼ 0 cx2 ¼ 0

3 Ex*3 ¼ Ex*0 þ PWaux þ PWb þ Ex*2 ¼ PWaux þ PWb cx3 ¼ PWauxþPWbqp$exðbp ;Tp;ppÞ

4 Ex*4 ¼ PWaux cx4 ¼ 15 Ex*5 ¼ PWb cx5 ¼ 1

Table 19Summary of the exergoeconomic parameters used in the analysis.

Exergoeconomic parameters

Parameter Value

Ratio of investment to equipmentcapital cost, invz

N/A

Number of years to built the plant, Pz N/ADistribution of investment costs along N/A

5.3. Retrofit No.5: isobaric energy recovery device as a BOP (boosterpump) and reuse of the Pelton turbine e hybrid retrofit

Another retrofit configuration is possible. It is only useful whenthe existing Pelton turbine is able to operate in a lower flow andpressure operation point in comparison with nominal operation.The new turbine flow and pressure working condition is supportedby the 2nd-stage brine rejected of retrofit No. 4. In Fig. 6, the layoutof the proposal is shown.

The energy obtained in the turbine increases the energy avail-able in the HPP motor and avoids installing another motor. Theretrofit is simplified but at the expense of producing less water inthe 5.000 m3/d train, the capacity of which is reduced. Themanoeuvres in this configuration are marked by the possibleretrofit of the Pelton turbine which recovers the energy from thebrine reject flow at a lower operation point. The possibilities ofdesign for the RO trains 1 and 2 are many and varied, and rangefrom decreasing the number of PVs to changing the train 1 recoveryrate to continue feeding all the pressure piping.

This configuration does not represent an improvement over thetwo previous retrofits and is only recommendable in such cases asthe installation of the new equipment or an increase in productionis limited due to lack of space.

Fig. 8. Flow chart of the whole productive process for the analysis in the case of ret-rofftting desalination plants (installation of energy recovery device based on isobaricchambers).

A hybrid retrofit was carried out in the Larnaca SWRO plant(Cyprus). In late 2008, an ERI ERD was installed and the productionwas increased from 54,000 m3/d to 64,000 m3/d. The SEC wasreduced by about 6% [13].

6. Methodology of the thermoeconomic analysis

Thermoeconomy is a useful and fruitful tool that combinesthermodynamics and economics criteria. Romero et al. (2005) [24]describes the thermoeconomic fundamentals for an SWRO plantbased on the Valero and Lozano approach (1994) [25]. Within thiscontext, the thermoeconomic analysis provides valuable informa-tion for each retrofit proposed about the influence of the efficiencyand cost of equipments on the efficiency of the whole plant, thecost of the product or the energy savings. Besides, this assessment

the years of built, xIzi (I, year)Taxes, Tz 0.35Percentage of capital cost to salvage

value, svzN/A

Amortization period, dz 8 yearsLifetime, Nz 15 yearsLoan interest rate, knz 0.06Mean inflation rate, hz 0.02Annual increasing of capital goods above

or below inflation rate, ecz0.00

Investment costs of the equipments, Zj(i ¼ 0: Intake ans pretreatment; i ¼ 1:pumping and ERD and RO rack)

Variable for each case (V)

Years in which the devices should bereplaced, mzlj

Intake, pumping and ERDonce during lifetimeMembranes: every five years

Specific O&M cost (insurance, labour andoverheads, breakdowns, fuel excluded), omzj

Variable for each case (V/m3

produced)Electricity cost, cW 0.12 V/kWhAnnual increasing of O&M costs above or

below inflation rate, mrz0.00

Annual increasing of product cost above orbelow inflation rate, evz

0.00

Annual availability, av 0.95

Table 20Input data costs in each retrofit case analysed.

Input data costs

Parameter RetrofitNo. 1

RetrofitNo. 2

RetrofitNo. 3

RetrofitNo. 4

Subsystems 0 cost,Z0 (Intake andpretreatment)

0.203 MV 0.203 MV 0.407 MV 0.435 MV

Subsystems 1 cost,Z12 (pumping,RO rack andERD if it isrequired)

1.132 MV 1.042 MV 1.310 MV 1.425 MV

Specific O&M cost,omzj (insurance,labour andoverheads,breakdowns,fuel excluded)

0.135 V/m3 0.135 V/m3 0.130 V/m3 0.140 V/m3

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626622

can identify where the major chances of improvements are in theproduction process in comparison with the standard case, i.e. theexisting SWRO plant. A whole thermoeconomic analysis includesexergy, exergy cost and exergoeconomic cost balances. Somerelevant references are: Usón and Valero (2009), Moran (1999),Bejan and Mamut (1999), Bejan et al. (1996) or Tsatsaronis (1993)[26e30].

The thermoeconomic assessment realized tries to compare thedifferent retrofits proposed:

� Retrofit No.1: The same HPP and installation of a new isobaricERD.

� Retrofit No.2: Isobaric ERD with a low pressure BOP to increasethe low feed water pressure.

� Retrofit No.3: Higher capacity per train and installation of anisobaric ERD.

� Retrofit No.4: Installation of an isobaric ERD used as a BOP.

ERetr. No.1

Retr. No.2

Retr. No.3

Retr. No.4

3.03

2.3

0.73

2.78

2.1

0.68

1.6

0.33

0.31

Fig. 9. Effective rate of fixed cost of the four retrofits analysed (15-year lifetime, 2% mean inflpumping, RO rack and ERD if it is required. Retrofit analysed: Retr. No.1: The same HPP andincrease the low feed water pressure, Retr. No.3: Higher capacity per train and installation

The exergy analysis is the first step of a thermoeconomic anal-ysis. The productive process is analysed for two subsystems andfive mass flows (j). Seawater intake and existing pretreatment(subsystem 0) and RO train and the pumping equipment e ERDin each case (subsystem 1). The exergy analysis performedcompares the exergy demanded by standard configuration incomparison with retrofits proposed. The system analysed hasdifferent inputs and outputs of exergy power flows, E

�

xj (kW) e seeFig. 7 in the case of existing plants and Fig. 8 for retrofitting cases -,considered as an open system at steady state conditions.

The purpose of subsystem 0 is to pump the raw water from thebeach well to HPP inlet. The pretreatment step is included within it.The purpose of subsystem 1 is to increase the seawater pressure tothe high pressure required, the RO process in the rack and recoverthe energy of the brine.

In this case, the only product, P, is the exergy power of thepermeate. Losses, L, is the exergy of the mass flow with no furtherusefulness e i.e. the blowdown of the RO train is processed by theERD of the SWRO plant. After that, the blowdown is an uselessoutlet flow.

Therefore, L (j¼ 2) is null. P (j¼ 3) is generated due to decreasingof the exergy of the pressurized feed as it circulates along the PV.The fuel, F, is the difference of the exergy power from the feedinput to the blowdown output. In this case the sum of the exergypowers of the seawater input (j¼ 0) and both subsystems (j¼ 4 andj ¼ 5 respectively.

According to the purpose equipments, the Table 17 shows the F,P and L of both equipments.

The rest of the useful parameters for the whole process arecalculated from P, L and F:

- The exergy power destroyed by irreversibilities, D: isa measurement of the irreversibility of the system analysed:

D ¼ F � P � L ðkWÞ

q. 0

Eq. 1

Total

1.93

2.01

1.7

0

0,5

1

1,5

2

2,5

3

3,5

Equipments

fo

etar

evitceffE

)s/€c(

stsoc

dexif

ation rate, 95% annual availability). Subsystems: Eq. (0) intake and pretreatment, Eq. (1)installation of a new isobaric ERD, Retr. No.2: Isobaric ERD with a low pressure BOP toof an isobaric ERD, Retr. No.4: Installation of an isobaric ERD used as a BOP.

Fig. 10. Exergoeconomic costs flow chart by each retrofit case analysed. Subsystems: Eq. (0) intake and pretreatment, Eq. (1) pumping, RO rack and ERD if it is required. Retrofitanalysed: Retr. No.1: The same HPP and installation of a new isobaric ERD, Retr. No.2: Isobaric ERD with a low pressure BOP to increase the low feed water pressure, Retr. No.3:Higher capacity per train and installation of an isobaric ERD, Retr. No.4: Installation of an isobaric ERD used as a BOP.

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626 623

- The inefficiencies, I: is the part of fuel which is not used inproduct generation. In this particular case I is equal to D due tothe null value of L.

I ¼ Lþ D ðkWÞ

- The exergetic performance, hx,: The efficiency assessment ofthe system from the point of view of the Second Law ofthermodynamics:

Fig. 11. Unitary exergoeconomic costs of the product of the system and each subsystem. Srequired. Retrofit analysed: Retr. No.1: The same HPP and installation of a new isobaric Epressure, Retr. No.3: Higher capacity per train and installation of an isobaric ERD, Retr. No.

hx ¼ PF

¼ 1� IF

ð%Þ

- Ratio of inefficiency, di (%), with respect to the total F ofequipment. Then, the subsystems with the highest values ofd should significantly improve the overall system.

The second step of the thermoeconomic analysis is the exergeticcost analysis. The exergetic cost of an exergy stream is the exergypower required for its generation. The symbol used to indicate the

ubsystems: Eq. (0) intake and pretreatment, Eq. (1) pumping, RO rack and ERD if it isRD, Retr. No.2: Isobaric ERD with a low pressure BOP to increase the low feed water4: Installation of an isobaric ERD used as a BOP.

Table 21Product cost for the retrofits proposed.

PARAMETER RetrofitNo. 1

RetrofitNo. 2

RetrofitNo. 3

RetrofitNo. 4

Product cost, cV/m3 54.2 53.1 56.7 66.0

Table

22Ex

ergy

analysis

resu

lts.

PARAMET

EREx

istingplant

RetrofitNo.

1RetrofitNo.

2RetrofitNo.

3RetrofitNo.

4

System

Eq.(0)

Eq.(1)

System

Eq.(0)

Eq.(1)

System

Eq.(0)

Eq.(1)

System

Eq.(0)

Eq.(1)

System

Eq.(0)

Eq.(1)

D�

103(kW

)(D

,%)

0.93

(100

%)

0.23

(24.7%

)0.70

(75.3%

)0.58

(100

%)

0.23

(39.7%

)0.35

(60.3%

)0.56

(100

%)

0.19

(33.9%

)0.37

(66.1%

)1.02

(100

%)

0.53

(52.0%

)0.48

(48%

)1.55

(100

%)

0.63

(40.6%

)0.92

(59.4%

)hx(%)

43.8

21.8

50.9

55.4

21.8

67.4

56.3

24.9

66.4

50.7

14.9

68.5

41.8

13.7

54.8

d i(%)

56.2

14.0

42.2

44.6

17.7

26.8

43.7

15.1

28.5

49.3

25.9

23.4

58.2

23.7

34.5

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626624

exergetic cost is a superscript, *. Then, exergetic cost of the product,P* is equals to F* for all subsystems.

The unitary exergetic cost of one flow, cxj, is the ratio of itsexergetic cost to its exergy power (dimensionless parameter). Inthe case of this system, Table 18 summaries the exergetic costs. Thefollowing assumptions have been considered:

- The exergetic cost of an input of the overall system is itsexergetic power.

- The unitary exergetic cost of the feed and the blowdown arethe same due to the fact that they are part of the fuel.

Once the exergetic cost balances are solved, the exergoeconomicanalysis completes the thermoeconomic analysis of this system. A15-year lifetime (Nz), a 2% mean inflation rate (hz) and an annualavailability (av) of the 95% have been considered for the retrofitproposed. Salvage values are not taken into account. For economiccalculations, Tables 19 and 20 summarize the exergoeconomic inputparameters used in the assessment. The rest of operation parame-ters needed (qvj, pj, bj, Tj, SECs) have obtained within the previoussections. Appendix A shows the respective exergoeconomic equa-tions for the system as a whole and for each subsystem.

7. Thermoeconomic results

Once the exergy powers and exergoeconomic input parametershave been considered, including the investment costs of differentretrofit cases, the unitary exergoeconomic costs, cP (V/kJ), of theproduct for the system as a whole and for every subsystem iscalculated. This unitary cost reflects the influence of irreversibilitiesand fixed costs in the generation of the useful products. For anyflow, the rate of exergoeconomic cost, Pj (V/s) is given by theproduct of its unitary exergoeconomic cost and its exergy rate.

Next figures compare different aspects for the four retrofitsanalysed. Fig. 9 shows the rate of fixed costs (V/s) for everysubsystem along 15-year lifetime. Fig. 10 represents the exer-goeconomic costs of every flow.

Feedwater (0) has zero cost because its generation does not needany process in this system. Blowdown (2) has a zero cost too since itis considered that brine rejection requires neither monetary costnor exergy consumption. In addition, no later usefulness is consid-ered. As it is showed in the cost of flow 5, the installation of theisobaric ERD increases the performance of the system and decreasesthe exergoeconomic cost of the product. Retr. No. 2 presents anadvantage in comparisonwith Retr. No.1 due to a lower cost of fuel.Retr. No. 4 presents worse product costs than Retr. No. 3.

The unitary cost of the product for every subsystem and retrofitcase study is presented in Fig. 11 and Table 21. Retr. No. 2 obtainsa slight better unitary exergoeconomic cost due to a lower value inthe subsystem 0. Retr. No. 3. obtains better unitary cost in bothequipments and in the overall system. Retr. No. 4 presents anincrease of 17% of total unitary cost in comparisonwith Retr. No. 3. Ingeneral, the highest differences of unitary costs are present in thesubsystem0. This increase is basedon the influenceof thefixed costsof the retrofitting in the case of change or increase the intake andpretreatment devices in comparison with the low exergy gained.

Retr. No. 2 and 3 present better results in comparison with Retr.No. 1 and 4 respectively. In the case of retrofitting an existing

Table 23Comparison of an SWRO plant with ERT with different retrofit configurations e The practical case of a 10000 m3/d SWRO.

Criteria SWRO plant withERT (Pelton)

Retrofit No.1HPP þ isobaric ERD

Retrofit No.2BOP þ HPP þ isobaric ERD

Retrofit No.3 HigherRO train þ isobaric ERD

Retrofit No.4 isobaricERD used as a BOP

Total capacity (m3/h) 2 � 5,000 10,000 10,000 2 � 7,200 2 � 5,000 þ 2 � 2.693No. vessels/No. Elements 104/728 104/728 104/728 150/1050 104/728e62/434Total feed water flow (m3/h) 926 926 926 1334 1436No. of HPPs pumps required 2 1 1 2 2No. of BOPs pumps required 0 1 2 1 0Civil work requirements n/a 5 5 \ \\

Requirement of additional area n/a 5 \ \\ \\

Hydraulic modifications n/a \\ \ \ \\

High voltage electrical modifications n/a 5 5 5 \

Low voltage electrical modifications n/a \ \\ \ \

Extra brine pipes installation n/a No No No YesEnergy requirement (kW) 1363.6 1014.2 1029.2 1424.08 1933.2SEC (kWh/m3) 3.27 2.43 2.47 2.39 3.02Capacity increase n/a 0% 0% 44.2% 53.8%Energy savings range n/a 25.6% 24.5% 0% 0%

n/a: not applicable.5 nothing; \ not too much; \\ relevant.

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626 625

installation for reducing the total energy consumption it should bemore adequate to use the same HPP preceded by a low pressureBOP and install a new isobaric ERD. In the case of retrofitting forproducing more freshwater it should be more adequate to increasethe capacity of the RO train and installation of a new isobaric ERDinstead of installing a 2nd-stage and using the ERD as a BOP for it.

The thermoeconomic difference betweenRetr. No. 3 andRetr. No.2 is minimal. In the case of installations without space limitations itshould be recommended to propose Retr. No. 3 instead of No. 2. Theincrease in more than 40% of the total capacity and the decrease ofSEC in 0.12 kWh/m3 are enough advantages to compensate the costrequired. However, Retr. No. 3 has a higher product cost. Theincrease of fixed costs does not influence the exergy rate.

To obtain more information and compare the different retrofitscases proposed with the standard case, the Table 22 represents therest of exergy results.

The highest exergy destruction is always presented in equip-ment 1 except for Retr. No. 3. Retr. No. 4 shows the highest exergydestruction. Retr. No. 2 and 3 present the better exergy perfor-mance and lowest ratio of inefficiency in comparisonwith Retr. No.1 and 4 respectively.

8. Conclusions

Reduction of the SEC is the field with the most specific technicalresearch focus and effort in SWRO plants. For existing installationswith energy recovery systems consisting Pelton turbines, the mostsignificant challenge is how to reduce energy costs. Highest effi-cient isobaric ERD are used in order to produce highest savings inenergy consumption in the desalination process and/or to increasethe freshwater capacity of the installations, by taking full advantageof the plant equipment.

Managers of existing SWRO plants are thus faced with thedelicate and difficult decision to opt for retrofitting of the ERT.

In this paper, different designs for optimising an SWRO desali-nation plant with ERT have been analysed. The high efficiencylevels of ERD makes it possible for various different retrofits to becarried out. Some of these help to reduce the power consumption ofthe installation and others increase the capacity significantly. Fiveretrofits are analysed under technical criteria and four of them arecompared under a thermoeconomic assessment.

The five retrofit described and calculated have been:

� Retrofit No.1: The same HPP is used and it is required to installan isobaric ERD in each of the two existing RO train.

� Retrofit No.2: Identical to previous retrofit but a low pressureBOP to increase the low feed water pressure before the HPP isneeded.

� Retrofit No.3: The same HPP is used but it is required theinstallation of an isobaric ERD in each existing RO train.

� Retrofit No.4: It is built a new RO train which will be the 2nd-stage of the existing RO train using the same HPP. The isobaricERD installed will run like a BOP/HPP of the 2nd-stage.

� Retrofit No.5: Identical to previous retrofit with the modifica-tion that the brine of 2nd-stage will pass through the existingPelton turbine.

Retrofit No. 1 and 2 are proposed to reduce the energyconsumption, whereas retrofit No. 3, 4 and 5 are proposed toincrease the total capacity of the existing SWRO plant. Tables 22and 23 concludes the main technical results of comparing anexisting 10,000 m3/d SWRO plant with the retrofits No. 1 to 4proposed.

A 15-year lifetime (Nz), a 2% mean inflation rate (hz) and anannual availability (av) of the 95% have been considered for theexergoeconomic analysis . The system studied is composed of twosubsystems: seawater intake and existing pretreatment (subsystem0) and RO train and the pumping equipments e ERD in each case(subsystem 1). The following conclusions are obtained:

� In general, the installation of an isobaric ERD increases theperformance of the system and decreases the exergoeconomiccost of the product.

� Retr. No. 2 and 3 present better results in comparisonwith Retr.No. 1 and 4 respectively.

� Retr. No. 2 presents an advantage in comparisonwith Retr. No.1due to a lowcost of fuel. Retr. No. 2 obtains a slight better unitaryexergoeconomic cost due to lowest value in the equipment 0.

� Retr. No. 4 presents worse product costs than Retr. No. 3. Retr.No. 3. obtains better unitary cost in both equipments and inthe system. Retr. No. 4 presents an increase of a 17% of totalunitary cost in comparison with Retr. No. 3.

� The thermoeconomic difference between Retr. No. 3 and Retr.No. 2 is minimal. In the case of installations without spacelimitations Retr. No. 3 should be recommended instead of No. 2.

Acknowledgements

The authors wish to give special thanks to Mr C. Santana forhis significant contributions.

B. Peñate, L. García-Rodríguez / Energy 36 (2011) 613e626626

Appendix A. Thermo-economic equations

Table A.1Summary of the exergoeconomic equations for the assessment of the system as a whole.

Exergo-economic equationsInvestment costs, V Iz :¼ Zretr

Zretr :¼ Z0þ Z1 ð0 : subsystem 0;1 : subsystem 1ÞSalvage value, V Sz: ¼ 0

Operation and maintenance, V/year OMz :¼ omz$volumeP

Nz

Actualized amortization, V Daz :¼ ðIzÞ � Szdz

$Xdz

t¼1

1ð1 ¼ knzÞt

Actualized investment costs, V Iaz: ¼ Iz

Actualized salvage value, V N/A

Actualized replacement, V

Raz :¼ Raz0þ Raz1þ Raz2

Raz0 :¼ Z0$h ð1þ hzÞð1þ knzÞ

imz01 þ Z0$h ð1þ hzÞð1þ knzÞ

imz02

Raz1 :¼ Z1$h ð1þ hzÞð1þ knzÞ

imz11 þ Z1$h ð1þ hzÞð1þ knzÞ

imz12

Raz2 :¼ Z2$h ð1þ hzÞð1þ knzÞ

imz21 þ Z2$h ð1þ hzÞð1þ knzÞ

imz22

Actualized O&M (excluding costs due to fuel consumption), V OMaz :¼ OMz$PNz

t¼1

hð1þmrz$ð1þ hzÞÞð1þ knzÞ

it

Actualized costs along the lifetime of the plant (excluding fuel consumption), V Zaz :¼ Iaz� Tz$Daz1� Tz

� Sazþ OMazþ Raz

Fuel, V/sPF :¼ P0 þP4 þP5ðflowsÞ

P0^^^^:¼ cP0$0 P4 :¼ cP4$PWaux P5 :¼ cP5$PWb

Fixed costs, V/s Zaz t :¼ Zaz½Nz$ð365$avÞ$24$60$60�

Product, V/s PP :¼ PF þ Zaz t

Unitary cost of productcPP kg :¼ PP

qpðV=kgÞ

cPP :¼ PPP

ðV=kJÞ

References

[1] IDA desalination yearbook. GWI Desaldata; 2009e2010. p. 6.[2] WoodcockDJ,WhiteM.TheapplicationofPelton type impulse turbines for energy

recovery on seawater reverse osmosis system. Desalination 1981;39:447e58.[3] Macharg JP. Retro-fitting existing SWRO systems with a new energy recovery

device. Desalination 2002;153:253e64.[4] Dundorf S, Macharg J, Seacord TF. Optimizing lower energy seawater desali-

nation, the affordable desalination collaboration. In: IDA World congress,Maspalomas, Gran Canaria, Spain; October 21e26, 2007.

[5] Miguel C, Cazurra T, Jurado JL, Orbe H, Sanz MA, Icaza G, Galtés J. Optimizaciónenergética de la desalinizadora del Llobregat mediante sistemas de inter-cambio de presión. Tecnología del Agua 2009;310:50e6.

[6] Stover RL. Retrofits to improve desalination plants. In: EDS world congress ondesalination and water reuse, Baden-Baden, Germany; May 17e20, 2009.

[7] Sadhwani JJ, Veza JM. Desalination and energy consumption in Canary Islands.Desalination 2008;221:143e50.

[8] Meerganz G, Medeazza V, Moreau V. Modelling of watereenergy systems. Thecase of desalination. Energy 2007;32:1024e31.

[9] FCCA - Fundación Centro Canario del Agua. Inventario de desaladoras enCanarias; 2006.

[10] Sánchez JM, Sánchez N, Sánchez R. Estado del arte en el diseño del proceso deplantas desaladoras de agua de mar de gran capacidad. Tecnología del Agua2008;301:64e9.

[11] Calder AG. Company web consulting, www.calder.ch; 28th October 2009.[12] Macharg JP. Energyrecovery technologyopensdoors tonewSWROplantdesigns.

In: American WaterWorks Association Conference, Reno, USA; October, 2002.[13] Stover RL. Seawater reverse osmosis with isobaric energy recovery devices.

Desalination 2007;203:168e75.[14] SchneiderB. Selection, operationandcontrolofaworkexchangerenergy recovery

system based on the Singapore Project. Desalination 2005;184:1177e90.[15] PeñateB,de laFuente JA, BarretoM.Operationof theROkinetic�energyrecovery

system: description and real experiences. Desalination 2010;252:179e85.[16] Bross S. SWRO core hydraulic system: extension of the SalTec DT to higher

flows and lower energy consumption. Desalination 2007;203:160e7.

[17] DANFOSS company website information, www.isave.danfoss.com; 27thDecember, 2009.

[18] Filmtec RO design software Rosa v.7.01. Dow Water Solutions Company;OctobereNovember 2009.

[19] Talo S, Suárez J, Bartels C, González M, Curbelo J, Abanades A, et al. Evolutionof configuration and operation regime at the Las Palmas III. In: IDA WorldCongress, Maspalomas, Spain; 2007.

[20] INALSA Company Technical Department - Personal communication, Mr. P.A.Martín and Mr. J.M. Bethencourt; November 2009.

[21] Energy Recovery Inc. product catalogues, www.energyrecovery.com;November 2009.

[22] Villa OL. Operating experiences of the Dhekelia seawater desalination plantusing an innovative energy recovery system. Desalination 2005;173:91e102.

[23] Hermony A, Pinto JM, Stover RL. Expansion retrofit of the via Maris Palmachimseawater reverse osmosis plant. In: IDAWorld Congress, Dubai UAE; November2009.

[24] Romero-Ternero V, Garcia-Rodriguez L, Gómez-Camacho C. Thermoeco-nomic analysis of a seawater reverse osmosis plant. Desalination 2005;181:43e59.

[25] Valero A, Lozano MA. CURSO DE TERMOECONOMIA. Spain: Department ofMechanical Engineering, University of Zaragoza; 1994.

[26] Usón S, Valero A. Analysis for the thermoeconomic diagnosis of energysystems. Int J Thermodyn 2009;12(2):61e7.

[27] Moran MJ. Fundamentals of exergy analysis and exergy-aided thermalsystems design. In: Bejan Adrian, Mamut Eden, editors. Thermodynamicoptimization of complex energy systems. Dordrecht, the Netherlands: KluwerAcademic Publishers; 1999. p. 73e92.

[28] Bejan A, Mamut E, editors. Thermodynamic optimization of complexenergy systems. Drdrecht, the Netherlands: Kluwer Academic Publishers;1999.

[29] Bejan A, Tsatsaronis G, Moran M. Thermal design and optimization. New york:J. Wiley and Sons, Inc.; 1996.

[30] Tsatsaronis G. Thermoeconomic analysis and optimization of energy systems.Prog Energy Combust Sci 1993;19:227e57.