THE ECONOMIC JUSTIFICATION OF INVESTING IN ENERGY EFFICIENCY

Transcription

1 Real (2015) world raw sugar price (US$/mt) REFEREED PAPER THE ECONOMIC JUSTIFICATION OF INVESTING IN ENERGY EFFICIENCY FIELD StJ AND SARIR EM Carbo Solutions International, Los Angeles, California, USA Abstract In some parts of the world, sugar factories have migrated from the historical model to the very latest model of high pressures and extremely low process steam usage. The historical model is characterised by low to medium boiler outlet steam pressure and high process steam usage, designed to consume all bagasse so as not to cause an environmental problem of disposal of surplus bagasse, without requiring the use of coal. The migration is expensive but is justified by the returns. The enablement of substantial cogen and other processes (because the energy situation has been optimised) represent a lifeline for an industry under pressure. In South Africa the migration has not taken place and there are still many historical factories. The various sugar companies have probably done their economic evaluations, but none have presented a paper at SASTA on such matters. This presentation aims at spreading the concepts to a wider audience. Introduction There are some very interesting trends available from the commodity brokers LMC 1 showing the global price of raw and white sugar in real terms. Brokers typically apply an exponential trend (dotted line) to the logged values of prices, which removes the impact of the short term (and infrequent) spikes in prices which are believed to provide a more reasonable long term trend (Figures 1 and 2) Real (2015) World Raw Sugar Price Exponential Trend Figure 1. Raw sugar global price trend in real terms. 1 Personal communication with John Adams, Senior Research Economist, LMC International, 4th Floor, Clarendon House, 52 Cornmarket Street, Oxford, OX1 3HJ. Tel: jadams@lmc.co.uk 160

2 Real (2015) world white sugar price (US$/mt) Real (2015) World White Sugar Price Exponential Trend Figure 2. Refined sugar global price trend in real terms. Data are not presented here showing the trends in domestic sugar prices in real terms, and it would be interesting to see them. However, the global trends alone are sufficient to indicate that the SA industry is under pressure, as a substantial proportion of the sugar produced in SA is sold on the international market. To be absolutely sure of such a statement, one would have to review the trends in real costs of production over the same timescale; but can it be assumed they have not reduced? If these data and thoughts are realistic, what can the South African (SA) industry do to alleviate the problem of diminishing returns and escalating costs? Can SA emulate the Mauritius and Thailand industries, to mention two, which have gone strongly into ethanol and cogeneration (cogen) in addition to their sugar manufacturing roles? There is a political component to the efforts that will be needed to transform the SA industry; however, this paper will look only at the technical issues. In SA there are 14 sugar factories operating at medium steam pressures, low cane throughputs and high steam-to-cane ratios. To represent them, imagine a factory processing tons of cane per day and then look at the options that may be exercised in terms of greater efficiency of energy utilisation. Assume that there is a revenue stream from the sale of cogen electricity of between 5 and 10 US cents/kwh. This stream will be examined, and the payback periods for the implementation of various new technologies will be calculated. The sugar revenue streams will be ignored, as they should continue as before in all cases studied. Of course the ideal would be to increase factory daily cane throughput capacities to levels such as to tons per day as in other countries. The word historical will be used often to describe plant and equipment that has largely not changed for decades and is characterised by medium boiler outlet steam pressure and high process steam requirements per ton of cane. 161

3 The hypothetical factory: raw + refinery Fixing of the bagasse parameters is important as the boiler performance data to be considered later need to form a consistent set (Table 1). Table 1. Input data for the hypothetical factory. Inputs % T/day T/h Cane Brix in cane Fibre in cane Moisture in cane Ash in cane 1.35 Bagasse Brix in bagasse 2.00 Fibre in bagasse Moisture in bagasse Ash in bagasse 1.50 Options to study There are two main variables, the rest being largely fixed for a given factory size: 1. Steam % cane (process steam requirements). 2. Boiler steam pressure and temperature. Turboalternator specific steam consumption (kg/h)/kw is not an independent variable, as it is related to boiler steam pressure and temperature. These give rise to four combinations that are of interest. Case Boiler pressure Steam % Cane Comment 1 High Low New factory design that maximises cogen. 2 High High New boilers and cogen but with process unchanged. 3 Medium Low Historical boilers and cogen but with process improvements. 4 Medium High Historical factory design - can implement very limited cogen. To study these options, it is useful to first look at typical turboalternator capabilities as the steam pressure (temperature follows suit) is increased. 162

4 The turboalternator 2 To clarify some terminology: there are two principal types, backpressure and condensing. Backpressure: the final, outlet steam passing to a process user is termed exhaust (typically 2.2 bara in an evaporator 1st effect). Condensing: the final, outlet steam passes to a non-process condenser (typically 0.12 bara). Both backpressure and condensing types can release steam at one or more intermediate points along the axis of the turbine, termed passout or extraction. This release of steam can be fixed or controlled (Table 2). Design power (MWe) Table 2. Turboalternator typical capabilities (backpressure). Inlet steam pressure Inlet steam temp ( C) Backpressure turbine inlet steam rate (kg/h) Exhaust steam pressure Exhaust steam temp ( C) Power efficiency (%) Sp. steam consumption (kg/kw.h) PG The two historical turboalternators at Pongola Mill are included in the table. A similar table of typical performance capabilities for a condensing turbine (with a small degree of extraction) is shown in Table 3. Design power (MWe) Inlet steam pressure Table 3. Turboalternator typical capabilities (condensing). Inlet steam temp ( C) Condensing turbine inlet steam rate (kg/h) Extraction pressure Extraction steam rate (kg/h) Condensor pressure Condensor steam temp ( C) Power efficiency (%) Sp. Steam consumption (kg/kw.h) When comparing the condensing turboalternator against the backpressure machine for the same steam conditions, the additional power obtained per unit steam rate is entirely due to the utilisation of most of the energy in the steam for electrical generation instead of some for electrical generation and some for evaporation of water in the first effect. These characteristics are trended below in Figures 3 and 4). 2 Personal communication with Ed Cliffe, Business Manager, Allen Steam Turbines, part of Weir Engineering Services Ltd, Highfield Parc, Highfield Road, Oakley, Bedfordshire, MK43 7TA, UK. Mobile/Cell: Direct Tel: Ed.Cliffe@weirgroup.com 163

5 Specific Steam Consumption, kg/kw.h Back Pressure Turbine Steam Flow Rate kg/h Condensing Turbine Steam Flow Rate kg/h Reduction in Turboalternator Steam Rates with Increase in Steam Pressure for the same Power Output MW 40 MW TA Inlet Steam Pressure, bara Back-pressure turbine Condensing turbine with extraction Figure 3. Steam rates to extraction and condensing turboalternators vs steam pressure. 7,5 Reduction in Turboalternator Specific Steam Consumption with Increase in Steam Pressure 7,0 6,5 6,0 5,5 Backpressure turboalternator 5,0 4,5 4,0 3,5 3,0 Condensing turboalternator with extraction TA Inlet Steam Pressure, bara Back-pressure turbine Condensing turbine with extraction Figure 4. Specific steam consumption for extraction and condensing turboalternators vs steam pressure. Inverting the parameter specific steam consumption for a turboalternator gives the parameter specific power output kw/(kg/h), which is the power produced per unit steam consumption (Figure 5). 164

6 % Change in Specific Power Output Specific Power Output kw/(kg/h) 0,28 0,26 Increase in Turboalternator Specific Power Output, kw/(kg/h), with Increase in Steam Pressure Condensing turboalternator with extraction 0,24 0,22 0,20 0,18 0,16 Backpressure turboalternator 0,14 0,12 0, TA Inlet Steam Pressure, bara Back-pressure turbine Condensing turbine with extraction Figure 5. Specific power output for extraction and condensing turboalternators vs steam pressure. Many of the boilers in SA operate at 32 bara and it is interesting to see the percentage increase in specific power output with increase in steam pressure within a discipline (backpressure or condensing) but also across disciplines (from backpressure to condensing); this is a substantial number and can be as much as nearly double (Figure 6) % Change in Specific Power Output, kw/(kg/h), with Increase in Steam Pressure - base case 32 bara Change from backpressure to condensing turboalternator Backpressure turboalternator Turbine Inlet Steam Pressure, bara Condensing turboalternator with extraction Back-pressure turbine Condensing turbine with extraction Sp. Power Output change (overall) Figure 6. Percentage increase in specific power output for extraction and condensing turboalternators vs steam pressure at 32 bara (base case). 165

7 % Increase in Boiler Enthalpy Rise (kj/kg) A way to improve the profit situation is to raise the boiler outlet steam pressure (so as to enable maximisation of power generated per unit of steam flow) and to reduce the process steam demand (so as to be able to increase the flow to a condensing turboalternator and correspondingly reduce the flow to the backpressure turboalternator). The boiler 3 It is of interest to note that boiler efficiency is measured from inlet to economiser and if heating takes place between the deaerator and the economiser, it is excluded from the energy efficiency balance. The other point of interest is the phenomenon of decreasing specific enthalpy of superheated steam as the pressure reaches very high levels (Table 4). Some boiler characteristics are: The boiler is a dumb servant to the turboalternator. It is an enabler for the turboalternator to generate more power per unit steam rate. The boiler efficiency is largely constant at all pressures. Table 4. Typical data for a range of boiler steam pressures and temperatures. Outlet steam pressure Bara Outlet steam temperature C Outlet steam rate Tph Outlet steam sp. enthalpy kj/kg BFW temp at deaerator exit C BFW temp to economiser C Number of HP heaters BFW sp.enthalpy (inlet to econ.) kj/kg Boiler efficiency (NCV) % Bagasse consumption rate Tph The trend below for the boiler is the % increase in specific enthalpy rise versus increase in steam pressure, with 32 bara being the base case (Figure 7). 9,0 % Increase in Boiler Enthalpy Rise with Increase in Boiler Pressure - base case 32 bara 8,0 7,0 6,0 5,0 4,0 3,0 2,0 1,0 0, Boiler Outlet Steam Pressure, bara % increase in boiler sp. enthalpy rise Figure 7. % increase in specific enthalpy rise across the boiler vs pressure. 3 Personal communication with Gaurav Rastogi, Isgec Heavy Engineering Limited. Noida , India. Tel: Mob : / gauravrastogi@isgec.co.in 166

8 The data were calculated from the deaerator to boiler steam outlet, and the design code s practice of heating the boiler feed water elsewhere but not accounting for it was ignored. It is interesting to see that the percentages for boilers are single-digit figures whilst the turboalternator percentages are in the range of 10 to 50 and 50 to 90% in round numbers (see Figure 6). The reason for this can be understood from the Rankine cycle diagrams in Figure 7. h1 h typ 2395 typ (h1-h4) 3234 typ m = steam rate in kg/h M = bagasse rate in kg/h boiler 85 bara 1 turbine 0,12 bara 2 Boiler Eff = 100*m*(h1-h4)/(M*NCV) TA Efficiency = 100*MWe/(m*(h1-h2)) Sp Steam Cons = m/mwe Sp Power Output = MWe/m alternator MWe condenser 0,12 bara (h1-h2) 1047 typ (h2-h3) 2188 typ h4 h3 208 typ 207 typ (h4-h3) 4 3 Figure 8. Rankine cycle for condensing turboalternator. If, for example, the scenario depicted above is viewed as a base case, then increasing the boiler steam outlet pressure (and temperature) will give rise to an additional delta-h to add to the boiler and turboalternator enthalpies of 3442 kj/kg. The percentage increase is this delta-h divided by 3234 kj/kg in the case of the boiler, and divided by 1047 kj/kg in the case of the turboalternator. The larger denominator will have its effect. The only further mental agility required is to understand that the turboalternator power output is the enthalpy drop across it multiplied by the efficiency, so it can be represented by the enthalpy drop. The Rankine cycle for a backpressure turboalternator is shown In Figure 9. h1 h typ 2695 typ (h1-h4) 2923 typ m = steam rate in kg/h M = bagasse rate in kg/h boiler 85 bara 1 turbine 2,2 bara 2 Boiler Eff = 100*m*(h1-h4)/(M*NCV) TA Efficiency = 100*MWe/(m*(h1-h2)) Sp Steam Cons = m/mwe Sp Power Output = MWe/m alternator MWe evaporator 2,2 bara (h1-h2) 747 typ (h2-h3) 2177 typ h4 h3 519 typ 518 typ (h4-h3) 4 3 Figure 9. Rankine cycle for backpressure turboalternator. 167

9 Modelling the scenarios Case 1 Case Boiler Pressure Steam % Cane 1 High Low Comment New factory design - maximises cogen with condensing/extraction turboalternator (CE) INPUTS T/day T/h Cane Bagasse REQUIREMENTS MWh/TC MWh/day MWh/h Factory power requirement 0, S%C T/day T/h Process steam requirement TURBOALTERNATOR BACK PRESSURE T/day T/h Turboalternator MWe 17 Turboalternator steam pressure bara 85 Turboalternator steam temp C 520 Turboalternator efficiency 95 Process steam requirement via letdown % Steam from desuperheating water 4 LP steam avail ex CE turbine 12 Steam required by/available from TA 85 Process steam required from TA 85 TURBOALTERNATOR CONDENSING/EXTRACTION T/day T/h Turboalternator MWe 34 Turboalternator steam pressure bara 85 Turboalternator steam temp C 520 Turboalternator efficiency 95 Turboalternator steam requirement 127 Turboalternator steam extracted 12 SUPPLY T/day T/h Boiler steam output Boiler outlet steam pressure bara 91 Boiler outlet steam temp C 525 Boiler efficiency (NCV) 89 Bagasse rate to boiler Bagasse surplus The factory power requirement is estimated at a typical value of 0.04 MW per ton of cane. This scenario is characterised by a steam % cane of 29% and a letdown rate of 16% on this. The total process steam requirement of 121 tph is met with letdown steam, desuperheating water, backpressure turboalternator exhaust and CE turbine extraction. High-pressure boilers are implemented, the cogen power is 34 MW and a surplus of bagasse is available. 168

10 Modelling the scenarios Case 2 Case Boiler Pressure Steam % Cane 2 High High Comment New boilers and cogen with CE turboalternator but with process unchanged INPUTS T/day T/h Cane Bagasse REQUIREMENTS MWh/TC MWh/day MWh/h Factory power requirement 0, S%C T/day T/h Process steam requirement TURBOALTERNATOR BACK PRESSURE T/day T/h Turboalternator MWe 28 Turboalternator steam pressure bara 85 Turboalternator steam temp C 520 Turboalternator efficiency 95 Process steam requirement via letdown % Steam from desuperheating water 10 LP steam avail ex CE turbine 6 Steam required by/available from TA 140 Process steam required from TA 141 TURBOALTERNATOR CONDENSING/EXTRACTION T/day T/h Turboalternator MWe 16 Turboalternator steam pressure bara 85 Turboalternator steam temp C 520 Turboalternator efficiency 95 Turboalternator steam requirement 60 Turboalternator steam extracted 6 SUPPLY T/day T/h Boiler steam output Boiler outlet steam pressure bara 91 Boiler outlet steam temp C 525 Boiler efficiency (NCV) 89 Bagasse rate to boiler Bagasse surplus 62 3 The steam % cane is assumed at 48% and the steam requirement of 200 tph is met with letdown steam (22%), desuperheating water, backpressure turboalternator exhaust and CE turbine extraction. High pressure boilers are implemented, the cogen power is made up of 11 MW from the backpressure and 16 MW from the CE turboalternators, and there is a bagasse surplus. 169

11 Case Boiler Pressure Steam % Cane Modelling the scenarios Case 3 Comment 3 Medium Low Historical boilers and cogen but with process improvements INPUTS T/day T/h Cane Bagasse REQUIREMENTS MWh/TC MWh/day MWh/h Factory power requirement 0, S%C T/day T/h Process steam requirement TURBOALTERNATOR BACK PRESSURE T/day T/h Turboalternator MWe 12 Turboalternator steam pressure bara 32 Turboalternator steam temp C 400 Turboalternator efficiency 95 Process steam requirement via letdown % Steam from desuperheating water 5 LP steam avail ex CE turbine 12 Steam required by/available from TA 84 Process steam required from TA 84 TURBOALTERNATOR CONDENSING/EXTRACTION T/day T/h Turboalternator MWe 26 Turboalternator steam pressure bara 32 Turboalternator steam temp C 400 Turboalternator efficiency 95 Turboalternator steam requirement 125 Turboalternator steam extracted 12 SUPPLY T/day T/h Boiler steam output Boiler outlet steam pressure bara 32 Boiler outlet steam temp C 400 Boiler efficiency (NCV) 89 Bagasse rate to boiler Bagasse surplus 86 4 The steam % cane is assumed at 29% and the steam requirement of 121 tph is met with letdown steam (17%), desuperheating water, backpressure turboalternator exhaust and CE turbine extraction. Low pressure boilers are retained, the CE turboalternator power is 26 MW but 5 MW has to be diverted to factory needs leaving 21 MW for cogen. There is a bagasse surplus. 170

12 Case Boiler Pressure Modelling the scenarios Case 4: No Cogen Steam % Cane Comment 4 Medium High Historical factory design no cogen INPUTS T/day T/h Cane Bagasse REQUIREMENTS MWh/TC MWh/day MWh/h Factory power requirement 0, S%C T/day T/h Process steam requirement TURBOALTERNATOR BACK PRESSURE T/day T/h Turboalternator MWe 17 Turboalternator steam pressure bara 32 Turboalternator steam temp C 380 Turboalternator efficiency 92 Process steam requirement via letdown % Steam from desuperheating water 12 LP steam avail ex CE turbine Steam required by/available from TA 139 Process steam required from TA 139 TURBOALTERNATOR CONDENSING/EXTRACTION T/day T/h Turboalternator MWe Turboalternator steam pressure bara Turboalternator steam temp C Turboalternator efficiency Turboalternator steam requirement Turboalternator steam extracted SUPPLY T/day T/h Boiler steam output Boiler outlet steam pressure bara 32 Boiler outlet steam temp C 400 Boiler efficiency (NCV) 80 Bagasse rate to boiler Bagasse surplus The necessary balance between high steam % cane and high letdown, and the turboalternator s poor specific steam consumption, results in no opportunity for cogen. 171

13 Case Boiler Pressure Modelling the scenarios Case 4: with Cogen Steam % Cane Comment 4 Medium High Historical factory design with limited cogen INPUTS T/day T/h Cane Bagasse REQUIREMENTS MWh/TC MWh/day MWh/h Factory power requirement 0, S%C T/day T/h Process steam requirement TURBOALTERNATOR BACK PRESSURE T/day T/h Turboalternator MWe 23 Turboalternator steam pressure bara 32 Turboalternator steam temp C 380 Turboalternator efficiency 92 Process steam requirement via letdown % 5 10 Steam from desuperheating water 2 LP steam avail ex CE turbine Steam required by/available from TA 189 Process steam required from TA 188 TURBOALTERNATOR CONDENSING/EXTRACTION T/day T/h Turboalternator MWe Turboalternator steam pressure bara Turboalternator steam temp C Turboalternator efficiency Turboalternator steam requirement Turboalternator steam extracted SUPPLY T/day T/h Boiler steam output Boiler outlet steam pressure bara 32 Boiler outlet steam temp C 400 Boiler efficiency (NCV) 80 Bagasse rate to boiler Bagasse surplus High steam % cane but a smaller letdown margin allows for a small amount of cogen, still with no CE turboalternator. The cogen amounts to 6 MW in this scenario. 172

14 Modelling the scenarios summary of Cases 1, 2, 3 The revenue streams vary from 6 to 20 million USD, depending on the contract price for the electricity supplied to the grid, the length of season and the various modifications that have been implemented in the factory (Table 5). Table 5. Revenue streams for Cases 1, 2, 3. Case Turboalternator MWe exported to grid Factory operation/annum, hours Turboalternator GWh USD/MWH for electricity sold to grid Revenue USD x 10⁶ USD/MWH for electricity sold to grid Revenue USD x 10⁶ These scenarios also assume that cogen is based on bagasse and is practised only during the season. This is practised in Mauritius, with a difference being that they supply to the grid during the off-season period also, but using coal as the fuel. There is no mass storage of bagasse in Mauritius for off-season generation. In Thailand however, there are massive stockpiles of bagasse for cogen as well as refinery operation over 11 months, but with milling operations lasting only six months. When the returns on investment are examined, this requires a knowledge of the installed cost of boilers and turboalternators. The information utilised below was obtained from suppliers. Table 6. Boiler costs (all inclusive, installed). Outlet steam pressure bara Outlet steam temperature C Outlet steam rate tph Boiler cost (all incl) x 10⁶ USD Cost of boiler/(steam TPH x pressure) USD Expressing the costs as a function of steam rate and pressure yielded a fairly uniform trend line, except for the point at 51 bara (Figure 10). 173

15 USD Cost of Boiler/(Steam TPH x Pressure) Boiler Pressure, bara Cost of Boiler/(Steam TPH x Pressure) Figure 10. Boiler installed costs per (steam rate x pressure) versus pressure. In a similar fashion, turboalternator costs were obtained from a well-known supplier. The few points obtained were extrapolated to cover the scenarios of interest. Installed costs are approximately independent of steam pressure (and temperature) and are related more to power output. The results are shown in Tables 7 and 8, and Figure 11. Table 7. Backpressure turboalternator costs (all inclusive, installed). Backpressure Turboalternator Estimated Power Total TA only installed costs Total MW USD x 10⁶ GBP x 10⁶ GBP x 10⁶ GBP x 10⁶ Table 8. CE turboalternator costs (all inclusive, installed). Condensing/Extraction Turboalternator Estimated Power Total TA only Condenser installed Total costs MW USD x 10⁶ GBP x 10⁶ GBP x 10⁶ GBP x 10⁶ GBP x 10⁶

16 USD x Installed Cost for Turboalternators MWe Backpressure Turboalternator Condensing Turboalternator Figure 11. Extrapolated turboalternator costs (installed). The Payback Periods This required an assumption to be made concerning the cost of modifications to the plant to reduce steam usage per ton of cane. This has been set at USD 20 million. The costs of the boiler and turboalternator investments are read off the extrapolated data. The capacities required of these assets are found in the case studies. In some scenarios, the historical equipment is retained, so no costs are shown in the investment funding requirements. The revenue streams are shown again to render assimilation of the data easier. Case 1 Case 2 Case 3 Turboalternator Mwe exported to grid Factory operation/annum. hours Turboalternator GWh exported to grid USD/MWH for electricity supplied to grid Revenue USD x 10⁶ USD/MWH for electricity supplied to grid Revenue USD x 10⁶ PAYBACK PERIOD CALCULATIONS New boiler steaming rate installed, t/h Boiler cost installed USD x 10⁶ New TA backpressure installed power. MW TA backpressure cost installed USD x 10⁶ New TA condensing installed power. MW TA condensing cost installed USD x 10⁶ S%C improvement costs USD x 10⁶ Total costs Payback period, years (high tariff) Payback period, years (low tariff)

17 The best payback period occurs in Case 2 at high tariff and for a 6000-hour season, which is the scenario where process remains unchanged and new high pressure boilers and turboalternators are installed. However, it does not have the added benefit of contributing the most to the annual revenue stream once the project costs have been covered, namely USD 16.4 million for the high tariff, or USD 8.2 million for the low tariff. The best annual contribution comes from Case 1 (new high pressure boilers and turboalternators and process improvements to steam % cane) at USD 20.4 million (high tariff) and USD 10.2 (low tariff). Conclusions The returns on investment look good and the SA sugar industry should be emulating its counterparts in other parts of the world. Is it the lack of clarity from government departments on cogen export tariffs that is delaying this vital transition? If so the industry should be laying siege to them. There are people who have done this in terms of trying to influence regulators and government, applying every ounce of energy they have to this, and who have not achieved anything. These efforts should continue and perhaps be escalated to a new level. If the migration to the global models does not occur, the SA sugar industry will most likely contract, with tragic consequences for persons affected. Acknowledgements Thanks are extended to Weir Engineering Services, ISGEC and Landell Mills Commodities for the information they provided which forms the basis for this paper. Thanks also to the CEO (Emmanuel Sarir) of Carbo Solutions International for co-authoring this paper and accepting to bear the associated drain on resources. 176