Drivepower efficiency: Selected industrial processes in a Swiss chemical firm

Energy Vol. 18, No. 6, pp. 627-639, 1993 Printed in Great Britain. All rights reserved

Copyright 0

0360-5442/93 $6.00 + 0.00 1993 Pergamon Press Ltd

DRIVEPOWER EFFICIENCY: SELECTED INDUSTRIAL PROCESSES IN A SWISS CHEMICAL FIRM GABRIEL MAMANE University of Geneva, Centre Universitaire d’Etude des Problbmes de I’Energie, chemin de Conches 4, Case postale 81, CH-1231 Conches, Switzerland (Received 18 August 1992)

Abstract-The drivepower of selected industrial processes has been investigated in a Swiss chemical firm which produces fragrances and specialty chemical products. We found that 42% of the electricity goes to drivepower (pumps, compressors, etc.). The energy uses are however varied: steam, water, mechanical cooling, compressed air. Drivepower is being

used in all these applications and its trend is on the increase. The energy savings of drivepower in these facilities and in the wastewater-treatment plant, all prevalent in the chemical industry, were estimated. A non-exhaustive estimate shows that out of 19,091 MWh of annual electric power, 1204 MWh or 6.3% can be saved without major changes in the equipment. A method which includes a gross analysis of the electric grid and commensense steps is described.

1. INTRODUCTION

According to studies carried out in Switzerland and other industrialized countries, a considerable potential for electricity (drivepower) savings exists in industry, in particular in the non-intensive energy industrial sector.’ Although preliminary estimates of this potential have been published, few detailed studies exist especially for Switzerland, extending from a detailed analysis to the realization of energy improvements and their measurements. The University Center for Energy Studies in Geneva has a long established collaboration with a middle-sized chemical firm in the Geneva region in the field of energy. A first energy balance based on data from 1986 showed that 16.5% of the annual energy consumed was electric at a price of 1.5 million SFr ($1 = SF1.43), representing 38.8% of the energy bill. By 1991, the percentage of electricity had gone up to 19.7% at a cost of 2.3 million SFr and represented 73 TJ out of a total of 371 TJ. This figure will increase due to the introduction of new production lines and the expansion of infrastructure units, such as a water-pumping station or the production of steam. Such plants are usually classified as non-intensive energy consumers. Is this a true statement? Can one check how much energy is needed to produce a fragrance whose composition is often a mixture of more than 200 chemical products? The answer is no, since a mixture of so many components is usually made at different times and at different temperatures. Although the cost of energy (oil, gas, electricity) represents a small part (5%) of the raw materials cost, energy is nevertheless important for the firm in its search for autonomy, while facing supply problems on the one hand and the environmental-protection issues on the other. The motivation of this work on selected industrial processes was to estimate the improvements, which are always possible, in the use of drivepower in a chemical plant. Its goal was to develop a method which could be applied in other sectors of the same firm and eventually in the entire chemical industry. The number of electrically driven motors in the plant is about 2200; 1835 consume from 0 to 4 kW, 299 from 4 to 15 kW, 28 from 15 to 30 kW, 50 from 30 to 100 kW, and 8 have a power requirement greater than 100 kW. We found that 42% of the electricity goes to drivepower (pumps, compressors, etc.). The electric power and energy used by the firm in the last few years are given in Table 1. In view of the number

of motors

involved

and the variety 627

of industrial

installations

which

628

GABRIELMAMANE Table 1. Electric energy and power for the plant in the last few years.

Year

Annual electric energy (MWh) Peak power (W)

1988

14650

3132

1989

16164

3492

1990

17743

3888

1991

20295

4176

have multiple purposes, it seemed logical to establish first electric load profiles of sectors for the whole plant. We have made measurements which we analyzed; the results are reported in Sec. 2. This analysis and the identification of different types of installations allowed us to proceed with the second phase, a detailed analysis of the wastewater-treatment plant, (see Sec. 3). Section 4 deals with an investigation of a distillation facility. In Sec. 5, we present the basic installations and facilities needed for doing chemistry. Section 6 contains an extrapolation of the electricity (drivepower) savings for the Swiss chemical industry to the future and a scheme of the method we elaborated in our case study for the Geneva region. often

2. PRELIMINARY

ANALYSIS

Electric measurements over 6 weeks allowed us to establish an inclusive consumption list for the different sectors of the plant. Because the number of measuring instruments is large, the measurements were made during three sequential periods of 2 weeks each. Cross-checking of three data sets allowed us to reconstruct the total plant consumption with an error of less than 1.7%, thereby enabling us to obtain correlated distributions for all of the consumers in the plant. 2. I. The Electric Grid The power is being supplied at MV (medium voltage) by three phases of 18 kV, which are distributed in an internal loop via 10 substations that transform the power to LV (low voltage) of three phases of 240 V. All substations are equipped with a capacitor-battery of 150-300 kvar (reactive volt-amp&e) for each transformer. These are engaged or disengaged as a function of the power factor of the consumers downstream in order to improve this ratio and decrease the losses of current. As shown in Fig. 1, in a 3-phase transmission system, one defines P = power in W, Q, = fundamental reactive power in var, Q, = harmonic reactive power in var, S, = apparent power in VA (volt-amp&e), and the following relations hold: P = UI, cos cp,

(1)

Q1 = WI sin (p,

(2)

&=P+z,

(3) P

h Q

S

Q1

Q

Sl W

Qn

Fig. 1. Diagram of vectors in a 3-phase electric transmission system.

Drivepower efficiency of selected industrial processes

629

s = jlm~,

(4)

Q=diSE,

(5)

Q,, = VU”(l’, + 1: + a . a),

(6) where I!/ is the efficient value of the voltage of one phase, ri the efficient value of the current of conductor i, and cp the phase lead of voltage U over the current Z,. For the electricity rates, the concept of reactive energy is defined by W, = Iz Q dt in varh (var hour), the active energy is defined by W = J: P dt in Wh (watt hour), and thus the power factor is introduced as cos @ = W/v-. The rates to be paid to the utility include a demand charge (see the definition of C below) of 5 SFr per kW. Whenever the mean power factor (averaged over a month) is above 0.85, a bonus is given by the electricity producer to the consumer. Similarly, a penalty occurs if the mean cos $J is below 0.85. The calculation is as follows: bonus( - ), penalty( + ) = f0.5 lcos Q,- 0.851 (A + B + C),

(7) where A = 0.128 SFr/kWh x monthly electricity power during peak periods (6-22 h), B = 0.007 SFr/kWh x monthly electricity power during off-peak periods (22-6 h), and C = demand charge = 5 SFr/kW x the maximal registered power integrated over 15 min during peak periods. As may be expected, such a tax rate encourages the plant operators to cut down their peak power and improve the efficiency of their own grid (cos $I- 1). 2.2. Results For each transformer of a substation, we obtained a load curve for the active and reactive powers and also a curve for the maximal and mean currents of one LV phase. Each curve consists of about 1300 points. Hence, there is the possibility to discern relatively fast events related to every sector of the plant. The other curves we obtained, cos @ and the distributions are calculated. 2.2.1. Power (i) Active-Figure 2 shows the total load curve of the plant. The average during the week is 2696 kW, 3192 kW during the day (8-17 h), 2338 kW during the night (21-5 h), and 674 kW for 3600 3200 2800 2400

1600

kvar

x

Date

time

Fig. 2. Total load curve of the plant (active and reactive powers). Measurements were done every 15 min over 2 weeks.

GABRIEL MAMANE

630

Compensation

steps

225 200 175 150 2 125 2

100 75 50 25 0

Reactive

power

Ideal

0

0

1

2

3

4

5

6

I

4

5

6

7

Day

0.98 0.96

t

0.92

0.86 0

1

Fig. 3. Com~nsat~o~

2

3

Day of reactive power and ideal calculated cos Q)in one substation.

631

Drivepower efficiency of selected industrial processes

the weekend (arbitrarily defined as 48 h from Friday midnight to Sunday midnight). The maximum recorded peak is 3528 kW. (ii) Reactive- Some substations have negative values (capacitive, kvar < kW) and some are reactive (kvar > kW). Correct compensation allows one to reduce the reactive power and improve the cos #, as shown in Fig. 3. This control and tune-up can easily be done at all of the substations. 2.2.2. cos $ The mean value is about 0.98. Some capacitor-batteries better than others.

compensate

for the reactive power

2.2.3. Currem The maximum and mean currents of each phase allow one to detect aberrant or irregular consumers which weigh heavily on the grid. One substation presented high peak currents (above 2300 A). These are due to a direct start of a large motor (>200 kW), as shown in Fig. 4. We have even observed reflections on the MV lines caused by such high currents in the LV lines. 2.2.4. The electricity distribution Figure 5 shows the contribution of the different substations to the total electric power during three representative periods of the week chosen arbitrarily. For each period, the total energy is indicated as follows: 27,354 kWh (8-17 h), 18,201 kWh (21-5 h), 38,358 kWh (Saturday 0 h-Monday 0 h). The units for water pumping (substation 175), steam production (substations 172.1 and 172.2), mechanical cooling and compressed air production (substation 230) account for one third of the total consumption. 2.3. Discussion It is very likely that peak currents which are caused by the operation of large motors that drive heavy pumps and compressors perturb the electric grid. This grid supplies power to more 16007

1400-

1200-

IOOO-

A 800-

600

Fig. 4. Maximum current at one substation. Measurements were done every 15 min over 2 weeks. The high currents are limited to 15OOA, which is the maximum meter range. The actual current is 2300 A.

GABRIEL MAMANE

632

DAY Wednesday 8 - 17 h

s em+mnment

q pduction i-J adminiswlion

6

0

70.1

70.2

51

54

67.1

67.2

210

2W

172.1 172.2 175

220

230

300

220

230

300

220

230

300

NIGHT Wednesday 21-5 h 18201 kWh

6

0

70.1

70.2

51

54

67.1 67.2

210

200

172.1 172.2 I75

WEEKEND Sat 0-Mon 14

Oh

38358 kWh

12 10 8 6

2 0

70.1

70.2

51

54

67.1

67.2 210

200

172.1 172.2 I75

Fig. 5. Detailed balance of the electric power of the plant during three periods of the week. The numbers on the horizontal axis refers to the substations.

and more control devices and computers whose need is for cleaner power. Only a THDA (Total Harmonic Distortion Analysis) of voltage and current can give an accurate answer to this question and help locate the perturbers. The total load curve (Fig. 2) reflects such peaks, whose maximal peak determines the monthly demand charge. This amount is reduced by about 15,000 SFr (of about 200,000 SFr total) every month, thanks to the mean cos r$ which is good (= 0.98) but which can be improved since a capacitor-battery exists in every substation, as mentioned above (see Fig. 3). From the electricity distribution of the plant (Fig. 5) one knows the precise weight of each sector in the plant during different periods of the week. An important point related to every electric grid is its reliability. Let us mention two aspects which are the power cuts and the emergency grid that must guarantee the supply to delicate connections. Table 2 shows a balance of power cuts and microcuts since 1986. The emergency grid is assured by a 560 kVA generator, essentially used for supply of the necessary units needed to cool down chemical reactors and prevent pollution in case of a general power outage. The measurements of intensities and powers over 6 weeks enabled us to reconstruct the total plant consumption at less than 1.7%, therefore allowing to obtain correlated distributions of all

Drivepower efficiency of selected industrial processes

633

Table 2. Electrid grid reliabiiity of the plant. -

-

- YeaUM icroculs x time(sec

1986

norecading

198

power

faikIreexternalto ptant

road&

accident

”

18 industrialsaviees netwark

indusuial sclvieesnetw#rk industrialsgviees (roadwork& i&trial

servicesnetwork

24 x O
81 x Oct.zO.6 1 x 0.87

1991

1 42 x t < 0.2 1 x 0.2<1<1

in the firm and detect the ones in sectors presenting an irregular load profile, To exemplify our approach, we have chosen the following cases which led to a more detailed investigation: (i) the curves of the substation 300, the w~tewater-treatment plant, show a rather irregular shape; (ii) the curve of the substation 67 shows frequeny starts of a large motor which is oversized and runs inefficiently; (iii) the curve of the substation 230 (Fig. 8) also shows frequent on-off cycling of equipment, the one of compressors for the production of compressed air; (iv) the units for water pumping, steam production, mechanical cooling production, and compressed air production which count for one-third of the total consumption. A new center for chemical production using 2ONl kVA is currently being introduced. In order to reduce the peak power of the plant which will exceed 6000 kW, the possibility to operate it in batch mode during the valleys of the total load curve (Fig. 2) has to be investigated. Only approximately 150 kVA of variable speed drives are installed in the plant. Variable speed answers the multiple demands of drivepower in the chemical industry. consumers

3. THE WASTEWATER-TREATMENT

PLANT

(WWTP)

The load curve of the WWTP was very irregular but served as a basis for ~mprehensive investigation. We established a detailed balance of its drivepower consisting of a total of 62

634

GABRIEL MAMANE

motors with a capacity up to 463 kW. Because of its operational mode, the WWTP is a continuous consumer of electricity. The primary consumers are the turbine-aeration systems (4 x 80 kW) which supply compressed air to biological sludge via diffuser tubes on which fine bubble caps are fixed, at the bottom of aeration basins. Study of the oxygen (air) supply and load curves obtained from current measurements, instantaneously and over long periods of time, allowed us to test different modes of operation of the turbines. The inherent difficulty in all biological treatment processes of this type is the management of (a) the demand for 02, which is a function of the wastewater load, and (b) the thickening and recycling of sludge. We now examine each of these points. Aeration consists of providing oxygen (air) to activated sludge* in two basins of 1500 m3 each. The principle is to maintain a high concentration of microorganisms which need O2 for energy and remove the organic matter present in the water to be purifled. Two turbines are used for each basin, each driven by a two-speed induction (asynchronous) motor of 54 and 80 kW. The power of the four motors is 63-69% of the total power installed at the WWTP, and accounts for 86% of its energy consumption. The weekly consumption of one aeration system (one basin) is about 30 MWh. Until 1990, one oxygen-transfer meter by basin was used for control of the motors, hence in automatic mode; this resulted in very frequent starts. In 1991, we recommended an operation in manual mode which consisted in running at low speeds at the end of the week when the sewage load was low, at low and high speeds at the beginning of the week when the load was medium, and at high speeds in the middle of the week when the load was high. This suggestion had the following results: (i) a marked drop in the number of starts of the four motors; (ii) less harmonics in the electric grid; (iii) a drop in motor losses and less need for maintenance.3’4 At the same time, in one basin, the number of diffusers were reduced by half, following an aeration calculation. The energy consumption dropped by 48%. For the other basin we noticed 8% increase in consumption due to plugging of the bubble caps which decreases their oxygen-transfer efficiency. In the beginning of 1992, drainage of the two aeration basins was done and we noticed that almost all bubble caps, 400 in number, of the first basin were in a good mechanical shape. For the second basin, the diffusers were cleaned and their number was reduced by half as well, using a configuration similar to the first basin. Thickening and recycling of sludge are essential for good functioning of any WWTP. In case they are managed correctly, the response of the plant is fast5 since biodegradation is done within reasonable times. The high wastewater loading rates to the plant are limiting factors of both the thickening and the recycling of sludge since flow rates are too big. Figure 7 is a schematic diagram of the WWTP. Thickening of sludge occurs in the clarifier by gravity and recycling of sludge is achieved using a pump. Figure 6 shows the current during the start phase of a motor at high speed that drives a turbine. A recording was done every msec during 20 sec. The commutation point of the YA (read y-delta) start at 5.5 set is well seen. In principle, for such a YA start, the rise in speed of the motor should proceed until the A current value matches the Y current value at commutation. Here it is not the case. Clearly, such high currents induce perturbations on the electric grid. 3.1. Discussion It certainly saves energy to operate aeration systems manually as a function of the wastewater load, instead of using automatic control by one single oxygen-transfer meter per basin as the pilot device. This change is needed only three to four times per week. Representing 63-69% of the total power and 86% of$he WWTP energy, the aeration systems are the weakest components in efforts to improve a WWTP. It goes without saying that drainage of the biological basins and control of the diffusers must be done regularly (once or twice a year). Air diffusers are then checked for failure, cleaned or changed if necessary. We have observed a drop in motor currents after such maintenance: 65 vs 70 A at low speed, 105 vs

Drivepower

efficiency

of selected industrial

635

processes

600 -,

A

-200

-600

-I

-800 I I

-1000

4

--r- ---;---T-.

_,...~~~._,7[1--

, 10

0123456189

11

12

I ---[---,-13

14

_-- -15

16

17

-r

18

19

set Fig. 6. Maximum

current

of a motor during the starting

phase at high speed of an aeration

the wastewater-treatment

system in

plant.

117 A at high speed. The estimated energy savings are 36%. One solution to the undersizing problem which often occurs some years after the design of most WWTPs is extension of the primary treatment capacity. Primary treatment prepares the wastewaters for biological treatment and consists of removal of solids by screening, which is followed by equalization that levels out the variation in flows and concentrations in a mixing basin. Next comes neutralization, where required, because streams of different pH partly neutralize each other when mixed. Another solution is treatment of the sewage at the source. In any case, more efficient use of drivepower and better biological efficiencies of the WWTP will result.

4.

A

DISTILLATION

FACILITY

AS

A

PRODUCTION

UNIT

The energy demand of this facility which consists of eight distillation units has been investigated. The drivepower installed in the largest unit is 25.4 kW, split into eight pumps (one WASTEWATER

CLEAN WATER

RECYCLED SLUDGE PlJMP Fig. 7. Schematic

diagram

of the wastewater-treatment

plant

636

GABRIEL MAMANE

kW

kva1

Date time Fig. 8. Load curve of one substation: active and reactive powers. Measurements min over 2 weeks.

were done every 15

of which is 10 kW) plus a motor for a stirrer of 1.9 kW. A distillation apparatus6 allows separation of the constituents of a mixture in a column in which ascending steam transfers part of its constituents to the falling liquid. Compounds of the mixture must have different volatilities; the falling liquid being enriched by the less volatile constituents. The vapour pressure of an ascending liquid increases with temperature; when this pressure is equal to the total pressure in the column, boiling occurs and the temperature at this pressure determines the boiling point. Since this point is a function of pressure, the temperature needed for distillation can be reduced if pressure is reduced, by means of a pump for example. This allows distillation of substances which break up before reaching the boiling point at normal pressure. A distillation column will fractionate properly as long as it is in equilibrium. This equilibrium takes time to reach (OS-l.5 h) and should not be destroyed by sudden operations of backward surge, of vacuum, or of heating power (achieved by steam). A distillation batch in this apparatus lasts about 15 h. A vacuum of about 1 mbar needed in this apparatus is technically delicate. In principle, it consists of maintaining the value of 1 mbar in a circuit which is not totally closed. Moreover, sometimes, a chemical product can break up spontaneously and suddenly create a poor vacuum which must be rapidly corrected to 1 mbar. This procedure demands a powerful pump that is ready instantly and equipped with permanently cooled Dewars to trap the gases,’ thus stabilizing the whole system and avoiding atmospheric pollution. Besides electricity, the other energy “uses” found in the distillation facility are: steam, used for heating of the liquid reservoir; water, for cooling of the condenser; brine, for lowtemperature cooling; and compressed air. Nitrogen is commonly used for handling and security measures. It goes without saying that good vacuum control of the apparatus during the entire distillation process is essential and would save on all the energy uses just mentioned. However, by operating the condenser at rather low temperatures a compromise with regard to the distillation theory is made in order to meet the chemical standards and especially the olfactory standard of the firm. Hence, the overall energy efficiency of the distillation process is somewhat sacrificed in the benefit of the olfactory standard of the firm’s product.

Drivepower 5.

THE

UNITS

FOR

efficiency

of selected

WATER PUMPING, COOLING AND

industrial

processes

STEAM PRODUCTION, COMPRESSED AIR

631 MECHANICAL

The electricity consumption of these four infrastructure units is about one-third of total. The installed drivepower is 1065 kW for water pumping, 395 kW for three steam production furnaces, 1197 kW for mechanical cooling production, and 518 kW for compressed air production. We have established a detailed balance of power and working hours for all the motors (105 total) in these units. The energy consumption of the 15 large drives of these units served as a good sample for suggesting possible improvements and calculating the associated costs for retrofit. For water pumping, a static Kramer drive (if the speed range of the motor is 50-100% of synchronous speed) or a static Scherbius drive (for speeds above synchronous speed)8 should be used instead of the rheostat system which consists of an external resistance that absorbs energy and is inefficient for speed control. These two techniques reduce slip losses that exist in large drives and are cheaper than frequency converters. At 50% of nominal speed, 45% of the energy does work and 55% represent losses. The estimated energy savings are 21% and the cost of electricity in the range 0.156-0.213 SFr/kWh. For steam production, frequency-controlled fans should replace the throttle valve (fan strainer),Y as already implemented in a new furnace. The motor for such fans could be half-size. The estimated energy savings are 11.4% and the cost of electricity 0.097 SFr/kWh. For mechanical cooling production, it is logical to couple three different networks of the plant and also produce local cooling whenever the need is sporadic and far from a production center. Moreover, increase of storage capacity would lower the frequent on-off cycling of compressors and cut down peak power during peak periods of the day. This choice would allow to produce mechanical cooling at low tariff during the night. For compressed air, two identical compressors should be coupled, thus reducing multiple starts and saving energy. Furthermore, by modifying and coupling the networks, an old unit could be phased out. No calculation was done to estimate the energy savings and associated cost of electricity but for some mechanical cooling and compressed air units, annual savings in kWh were obtained. A totally centralized production for mechanical cooling, by means of a large d.c. (synchronous) motor, can be used to correct the overall power factor (cos c$) of the electric grid since the d.c. motor can generate the reactive power absorbed by the a.c. motors of the whole plant (see Ref. 8). Lastly, cogeneration is today cost-effective for units of 100 kW and more” and should be considered since both heat and drivepower are prevalent in chemical plants.

6.

AN

EXTRAPOLATION

FOR

THE

SWISS

CHEMICAL

INDUSTRY

TO

THE

FUTURE

The rational use of energy is necessary to achieve the goals of the programme Energy 2000 launched by the Swiss government following the acceptance, by referendum, of the constitutional amendment on energy of September 1990. It is also the national answer to various problems and conjectures on the future of our planet, such as the greenhouse effect, waste storage, energy supply, and protection of the environment. See e.g. Schmidheiny (Ref. 11). The Swiss chemical industry produces specific chemical agents not exceeding a few metric tons. It produces practically none of the so-called “mass-produced chemicals”, whose manufactured volume usually attains many thousands or even hundreds of thousands of metric tons per year. In 1980, the chemical industry consumed 25,296TJ, of which 6557 TJ of electricity or 26%. In 1990 the electricity reached 33.2% for an unchanged consumption of 25,405 TJ which account for 3.3% of the total energy used in Switzerland. This growth is moderate and indicates that the Swiss chemical industrial sector is not energy intensive in spite of large-scale introduction of new installations for the protection of the environment (wastewater-treatment plants, gaseous waste treatments, etc.). The part for drivepower is 53.1% of the electric energy and shows a small growing trend (see Fig. 9). In order to make

GABRIEL MAMANE

638

100% = 8372 TJ

90 80 r

0 total 70

E? drivepower

60

n steam production

50

q lighting

40

0 heating

30

0 self production •Ienvironment

20 10

..i 1986

1987

1988

0 1989

Fig. 9. Annual electric power consumed by the Swiss chemical industry. Consommateurs d’Energie de I’industrie et des autres branches

Source: Union Cconomiques.

suisse

des

substantial savings in electricity, it is logical to first investigate the drivepower sector which is the largest consumer. Based on our case study, it appears that the proposed method, often easy to carry out but difficult to have accepted by the concerned circles, is the real answer to the problem. Our working hypothesis is that all chemical plants present a similar electric load profile since chemical processes demand supplies of water, steam, cooling, and compressed air. Drivepower is present in all these energy uses; in water pumping, essential to chemistry and in large quantities, via the pumping station usually located on a river bank; in steam production, mainly for pumping water and fuel and for the fan blowers of furnaces; in mechanical cooling and compressed air, usually for driving compressors. Drivepower is also omnipresent in the transport of raw materials and solvents since every pump is driven by an electric motor. Finally, the protection of the environment consisting of in-plant treatment or biological treatment of wastewaters is also a continuous consumer of drivepower. 6.1. The method Based on our study for a firm in the Geneva region, it is necessary to make a gross analysis of the infrastructure units for water pumping, steam production, mechanical cooling and compressed air, where large motors are found. This is the first step. A preliminary analysis of the electric grid which includes measurements of power, voltages, and currents for all the transformers (substations) in the plant in order to establish their load profiles, is the second step. Investigation of these curves allows one to obtain an overall view of the plant consumption and detect problems in sectors presenting an irregular or suspicious load profile. Questions raised in the preliminary analysis lead to a check downstream of the transformers, namely, to establish a detailed balance of drivepower in the problematic sectors. The large motors (>50 kW) are usually the source of abnormal consumption and to remedy this, the three following steps should be carried out: (i) verify the motor control (does the pilot device that controls the motor work properly?) and estimate an alternative mode which reduces the number of motor starts; (ii) verify the motor sizing if necessary and replace it by an energy-efficient motor; (iii) check all related equipment connected to the drive, a refinement of all components should be considered if one wishes to obtain optimal use of power supplied by the motor. This last point touches upon the key issue of any energy-conservation implementation, that is, rationalization of (drivepower) use. Although energy savings can be achieved by acting

Drivepower

efficiency

of selected industrial

processes

6.111

directly on the motor (e.g., exchange for a high-efficiency motor, use adjustable speed drive). optimization of the entire sequence is the real answer to the problem. Another point related to the use of drivepower is its influence on the quality of the electric grid. Peak currents due to starts of large motors are reflected on the grid which supplies more and more control devrces and computers that need clean power. Motors operated correctly induce minimal perturbations on the grid.‘* The method suggested here is a synthesis of a preliminary analysis of the whole electric grid of the plant plus a detailed analysis of its drivepower. Our case study in a specialty and chemical products plant has shown that energy savings in drivepower can often be achieved without major changes in the equipment. We assume that such potential exists in every chemical plant and to look into the problem of losses is the first step to cross, by the engineers of the plants first of all. A non-exhaustive estimate shows that out of a total consumption of 19,091 MWh, 1204 MWh or 6.3% can be saved in the plant. An extrapolation for the Swiss chemical industry of 6.3% is certainly conservative and would lead to energy savings of 2X3TJ (79 GWh). Indeed, an effort in this field seems justified. To do so, one needs to disseminate this method, establish relations with engineering departments of plants and appeal for their collaboration. It is through this procedure that electricity savings in the chemical industry can be assessed and prompt firms to take action to increase energy efficiency, We have noticed the following points: variable speed answers the multiple demands of drivepower in the chemical industry. Oversizing of motors is a known phenomenon; downsizing is worthy and should be examined after a running period of any drivepower application. it is generally beneficial to couple the various networks, such as mechanical cooling or compressed air, thus reducing multiple starts and phasing out old units. Their local production is usually cost-effective for remote workshops. Increase of storage capacity would lower frequent on-off cycling of compressors, cut down peak power and reduce costs, e.g., by producing cooling at night. Often, manual operation of motor systems instead of automatic control done by a poor pilot device is a good action to override design mistakes and/or unreliable response of instruments. In applications where a vacuum is required, its mastery is essential, savings on all energies of the process will follow. It is in the operating mode (management) of the different technical installations that the answer to energy saving is found. Being cost-effective for units of 100 kW and more, cogeneration is adequate for chemical plants. Acknowledgement-Financial acknowledged.

support

from

the

Swiss Office

F&d&al

des Questions

Conjoncturcllcs

is grarcfully

REFERENCES

1. A. B. Lovins, “The State of ihe Art: Drivepower”, Technical Report from COMPETITEK. Rocky Mountain Institute, Snowmass, CO (1989). 2. Mkmento Technique de I’Eau, Degremont, Paris (1989). 3. B. J. Chalmers, Electric Motor Handbook, Butterworths, London (1988). 4. Motor Application and Maintenance Handbook, R. W. Smeaton ed.. McGraw-Hill, New York. NY (1987). 5. W. W. Eckenfelder Jr., Ind~tr~al Water Pollution Control, McGraw-HiIl, New York, NY (1989). 6. A. F. Waterland, in Recycling, Fuel and Resources Recovery, pp. 7-31, M. Grayson ed., Wiley. New York, NY (1984). 7. I. I. Ionel, Pumps and Pumping, Eisevier, Amsterdam (1986). 8. S. Nadel, M. Shepard, S. Greenberg, G. Katz, and A. T. de Almeida, Energy-f$cicanr Motor Systems, American Council for an Energy-Efficient Economy ed., Washington, DC (1991). 9. B. Eck, Fans, Pergamon Press, Oxford (1973). 10. N. I. Meyer and J. S. Norgard, in Electricity Eficient End-Use and New Generation Technologies and Their Planning Implications, pp. 787-810, T. B. Johannson, B. Bodlung, and R. I-I. Williams eds., Lund University Press, Lund (1989). 11. S. Schmidheiny, Changing Course, MIT Press, Cambridge, MA (1992). 12. A. T. de Ahneida, S. Greenberg, and C. Blumstein, IEEE Trans. Power Sys. 5, 852 (1990).