Section 4.

Optimizing Energy Usage

Beginning with Conclusions

After only a brief study of Figure 23, it appears that the optimum battery selection is one that results in 80% discharge by the end of a pre-established work program. Said another way, the battery should have a rated capacity of 125% of the energy it is expected to deliver during discharge. It makes no difference whether the intention is to charge the battery at the end of each workshift or whether the battery is to be charged at the completion of a particular task. The conclusion remains the same: the battery should be placed on charge when it has been 80% discharged.

Selecting the Correct Battery Capacity

The selection of battery capacity for a given truck and application becomes more significant with each increase in the capital cost of batteries and charging stations as well as with each increase in the cost of energy.

Each industrial truck application presents a separate battery-selection problem with more involved than the size or lifting ability of the truck. Since trucks may be used for different purposes, each application presents a specific work profile that can be thought of as a series of rapidly varying current drains.

In any application, this battery drain varies from instant to instant during the entire time the truck is in use. As shown in Figure 24, every task the truck performs represents a different battery drain ... that can, for example, range from 5 amperes, while steering, to 1000 amperes, motor in-rush current on the pump motor.

Figure 25 shows a hypothetical operation for a typical truck-lifting loads and transferring them to nearby locations. Our typical truck takes 6 separate steps to complete these operations, and each step drains energy from the battery.

Of course, there is no simple way to calculate in advance exactly how much energy will be needed to perform any single step of an operation, let alone a whole day's work. The only real solution is to measure the number of ampere-hours used by the truck when it performs each step. This is done with an ampere-hour meter installed on the truck. *

With the Power Prover Ampere-Hour Meter installed on the test truck, the driver performs the stipulated sequence of steps and the meter readings show the total ampere-hours and the amperes used in the operation. The number used in each step of the operation can be monitored by recording the meter readings while the truck operates.

 

*The Curtis Model 1020 Power Prover Ampere-Hour Meter is an accurate, easy-to-install ampere-hour meter widely used for this purpose.

 

A procedure of this kind that includes representative operations performed by the truck provides a simple and accurate basis for selecting battery capacity ... or for verifying assumptions about ampere-hour requirements.

Ways to Measure State-of-Charge on the Fork Lift Truck

The reason for measuring the state-of-charge as the battery is in use is twofold: to protect the battery from deeply discharging and, thereby, internal damage; and to protect the truck's electrical components from the negative effects of low voltages, a consequence of deeply discharged batteries. At average discharge rates of 8 hours or less, a measurement which accounts for the remaining capacity as a function of discharge rate can serve to protect both the battery and the truck.

An ampere-hour meter provides useful information that greatly simplifies specifying the correct battery and measures ampere-hours used, but not the rate at which it is used. And the rate at which it is used is crucial in measuring the state-of-charge because battery capacity is different at different rates of discharge.

Aside from ampere-hour metering, three basic measures have been used in industry to determine battery state-of-charge: specific gravity, open circuit voltage, voltage under load.

Specific Gravity: A Static Measure

Not only are specific gravity measurements not convenient to make during the work period, but their value is limited because it takes time for the specific gravity to stabilize after the battery load is disconnected. Although a convenient measure of overall battery condition, specific gravity measurements give no valid indication of the discharge history that produced the reading. Any given reading of stabilized specific gravity can either be the result of heavy discharge for a short period or of prolonged discharge at a very light load. This effect is shown in Figure 26, in which the specific gravity and the open circuit are plotted against % of discharge for currents from 25 amperes to 800 amperes for a 1200 ampere-hour battery rated at 6 hours. Any specific gravity line, for example, 1165 SG, intersects a number of discharge lines. 1l65SG, in particular, corresponds to 80% discharge at the 6-hour rate (200 amperes). However, if actual operation is at 600 amperes, the truck motor will have been subjected to repeated, excessively low voltages (less than 1.7 volts per cell) for a significant amount of time.

Of course, if the truck has been used in essentially the same manner during each workshift, then the specific gravity (unstabilized) measured at the end of the shift will be pretty much the same from day-to-day. If there were any sudden change in the reading, it would be informative, but would not reveal anything about the state-of-charge except in a general way.

 

 

Figure 24: Relative Current Drain for Typical Industrial Trucks

 

Figure 25: A Hypothetical Industrial Truck Operation

 

 

Open Circuit Voltage: Another Static Measure

Since specific gravity and open circuit voltage are directly related, a similar line of reasoning shows that unstabilized open circuit voltage is also not a valid measure of battery condition. In Figure 26, any of the constant voltage lines can represent any number of battery discharge histories. Hence the open circuit voltage is not a useful measure of state-of-charge during operations. Note that battery manufacturers always determine capacity by measuring voltage while the load is connected. Disconnecting the load and immediately measuring open circuit voltage reveals nothing about the state-of-charge.

Voltage Under Load: A Dynamic Measure

When the truck operates, it presents varying electrical loads to the battery. As soon as the battery is "loaded" the open circuit voltage drops abruptly to the initial value of voltage under load. As long as the load current stays constant, the voltage under load slowly decreases as the battery discharges. If we keep track of the average voltage under load, we can tell how fast the battery is being discharged at any instant. Voltage under load measurements account for the rate at which a battery is discharged, whereas specific gravity and open circuit voltage reveal nothing about rate.

Figures 27 and 28 illustrate aspects of voltage under load. In Figure 27, voltage under load is shown as a measure of state-of-charge for 5 constant currents from 0-100% discharged. In Figure 28, varying current rates, based on the work procedure illustrated in Figure 24, are shown as a magnified micro-section tracked by an instrument with appropriate electronic computing circuitry.

 

Figure 26: Stabilized SG and Open Circuit Voltage as Measures of State-of-Charge for Various Discharge Rates

 

 

 

Since time always moves from left to right in Figure 27, the net effect of many different loads is to move the measurement point steadily toward the right, always along one load line or another. On the fork lift truck, there are many more load line variations. Every interval, during which the measured voltage follows a given load line, contributes a particular percentage of discharge.

One way to accomplish voltage averaging in an instrument is to continuously compare the varying battery voltage with a reference voltage which changes as a function of battery state-of-charge. Whenever the battery voltage is less than the reference voltage, the time below the reference voltage is measured and stored in the instrument's memory. The output of the memory sets the value of the reference voltage and represents the state-of-charge of the battery. It is displayed on a meter ("fuel" gage) located right on the truck in the driver's view.

Figure 28 shows how the reference moves in response to the battery voltage and how the "fuel" gage displays the state-of-charge.

State-of-Charge-Based Charging Can Save Energy

Some plant managers feel that trucks must be available without interruption throughout a workshift. Thus it becomes necessary to provide each fork lift truck with a fully charged battery at the start of each shift and to return the battery for charging at shift end. In any two- or three-shift operation, this practice normally requires at least twice as many batteries as trucks plus a charging station for each truck. For substantial fleets this means large capital investments, considerable floor space, significant personnel requirements, and additional energy costs.

This practice of workshift-based charging is especially prevalent in manufacturing plants where stopping the assembly line for any reason cannot be tolerated. At one automotive manufacturing site, at which a stalled line would be excessively costly, management has initiated a program in which fully charged batteries are installed on 150 trucks in 40 minutes at the end of every workshift.

 

Figure 27: Voltage Under Load as the Measure of State-of-Charge

 

Figure 28: A Practical Application of Voltage Under Load as a Measure of State-of-Charge

 

 

For other than such above noted critical requirements, with the use of a reliable and economical means of monitoring battery state-of-charge on the truck, another approach to charge scheduling has emerged. It is no longer necessary to provide each truck with a fully charged battery at the start of each shift and to work the battery through to shift end. Rather, if there is still significant charge left in the battery at the end of the shift, the battery can be left on the truck and worked until the need for charging is indicated. Then, and only then, the truck returns for a freshly charged battery. Its discharged battery is then placed on charge and, 8 hours later, is ready for use again.

As shown earlier, if an 8-hour charge period is used (as is generally the case), the optimum discharge point is 80%. Figure 29 shows how rapidly the energy requirements rise when batteries are charged for 8 hours after having been discharged less than 80%. Working each battery to the 80% discharge point before returning it for charging permits the most effective use of energy in the system and provides significant savings in energy costs.

Since each of the batteries reaches the 80% discharge point at a time that depends on the way it is worked, most will work longer than one 8-hour shift. Since a new battery isn't required for each truck at the end of every shift, it isn't necessary to have at least one replacement battery per truck, nor is it necessary to have at least one charger for each truck. Fewer batteries and chargers mean less capital investment, less space used for charging, and fewer people to do the work.

 

Figure 29: How Energy Requirements are Affected by Depth of Discharge

 

Further, when a battery is placed on charge it draws maximum current from the line. Placing all of the fleet's batteries on charge at one time, therefore, creates a large demand, which is reflected in the cost of energy in the form of higher utility peak demand charges. However, starting batteries on charge at different times throughout the entire shift reduces this demand and, therefore, the cost of energy.

How A "Typical" Fleet Can Save Energy

To dramatize the energy saved by charging batteries under optimum conditions, we have prepared data for a hypothetical 20-truck fleet operating for a 5-day week of 2 shifts per day. The data cover one month of 20 working days and show the following:

        That it is possible to significantly reduce the amount of energy required to operate the fleet

        That by appropriate fleet management it is possible to greatly reduce the magnitude of peak power demand

        That this modified fleet operation holds the promise of reducing the capital and labor costs associated with industrial trucks.

Our data and calculations are not derived from operating a real fleet. They do, however, suggest how to reduce the cost of operation of any real fleet of trucks.

In this fleet, all trucks are equipped with the same battery type: a 36 volt, 1200 ampere-hour unit. In the original operating mode, each truck started each shift with a fully charged battery. We have divided the batteries into six classes (A-F) based on their average state-of-charge at the end of a typical shift. The classes, the number of batteries in each class, and the discharge data are listed in Section I of Table 4.

Section 2 of Table 4 shows the energy used by each battery (kilowatt-hours are calculated at the average output voltage). Section 3 then totals these per-battery figures by battery class. The overall efficiency is calculated by dividing the output energy (kWh) by the AC input energy and multiplying by 100. Since there are 40 shifts in our 20-day month, the totals are multiplied by 40 to obtain an estimate of energy used by the entire fleet over a full month of operation. A very substantial 31.6 megawatt-hours is used ... but at only 55% overall efficiency to produce the required 17.4 megawatt-hours of work delivered by the batteries.

 

 

Table 4: Energy Usage Data for a Hypothetical

20-Truck, 2-Shift Fleet

 

SECTION 1

Fleet Composition

SECTION 2

Battery Data

SECTION 3

Fleet/Shift Data

 

Class

&

Number

of

Trucks

 

Average

%

D.O.D.

at Shift

End

Output to

Load

Power

Line

Input to

Charger

(kWh)

Output to

Load

Power

Line

Input to

Charger

(kWh)

Overall

Efficiency

(%)

 

 

 

Ah

kWh

Ah

kWh

 

 

 

A-1

B-2

C-3

D-8

E-4

F-2

80.0

69.3

58.7

48.0

42.7

37.3

960

832

704

576

512

448

33.6

29.2

24.9

20.5

18.2

16.0

50.2

46.6

42.8

38.5

36.0

33,3

960

1664

2112

4608

2048

896

33.6

58.4

74.7

164.0

72.8

32.0

50.2

93.2

128.4

308.0

144.0

66.6

67

63

58

53

51

48

 

 

Fleet Totals Per Shift

Per Month (40 shifts)

12,288

435.5

790.4

31,616

55

55

 

 

491,520

17,420

 

 

By the simple expedient of returning each truck for a fresh battery when its "fuel" gage reads between 75% and 80% discharged, we can sharply reduce this energy waste. Since overall system efficiency is 67% for batteries that are charged after being 80% discharged, we can reduce the power line energy needed for our hypothetical fleet to 25.9 megawatt-hours if each battery is 80% discharged before being recharged. This means that about 5.7 megawatt-hours can be saved each month.

Reducing Peak Power Demand Costs*

If the 80% discharge point for our typical battery is 960 ampere-hours, then when that battery is discharged to 832 ampere-hours, or 69.3% (as in Class B), there is still a fraction of a workshift left in the battery. In fact, a Class B battery will actually last 1.15 shifts. Each of the other classes of battery will last correspondingly longer into the second shift, as shown in Table 5.

 

Table 5: Battery Workshift Capacity

Class

Ah Used

in One Shift

Total Ah

Available

Total Work shift

Capacity per Truck

per Charge (Shifts)

A

B

C

D

E

F

960

832

704

576

512

448

960

960

960

960

960

960

1.0

1.15

1.36

1.67

1.88

2.14

 

Table 5 shows how battery charges can be spread out across the two workshifts because individual batteries will reach 80% discharge at different times, depending on individual work profiles. With time, the spread will grow even more random, so that charge starts will be evenly spread throughout the two shifts. As shown in Figure 30, the net effect is to reduce the peak power demand below its absolute maximum of nearly 220 kilowatts. (The maximum peak power demand occurs when all 20 batteries are placed on charge at once.) For example, splitting the batteries into two groups of 10 cuts the peak demand to about 198 kilowatts, a 10% reduction. And in the optimum case, in which one battery starts on charge every 24 minutes throughout the shift, peak demand falls to 130 kilowatts, a reduction of over 40%.

 

 

*Peak Power Demand: The use of electricity is made up of varying periods of high and low demand. Power companies build and provide sufficient generating capacity to meet periods of highest demand. To help defray the large capital outlay made to build and maintain maximum generating capacity the electric company charges customers what is referred to as a "demand charge". The charge is levied as an added cost of electricity. Two meters are used to monitor the consumption of electricity by commercial and industrial customers: one to track kWh's and one to monitor highest average kilowatts of demand, The demand meter reads 15 minute intervals of which the two highest are used to make up an average half-hour for the monthly demand charge. Thus, it behooves a customer to limit concentration of demand for power. In the electric fork lift truck environment, this means not putting all batteries on charge al the same time or charging batteries at periods of low demand (nighttime).

 

Figure 31 is a more general chart that shows the effect of different charging schedules on peak demand for any size fleet. To use the chart, calculate the maximum peak demand  the peak demand when 100% of the batteries are placed on charge at the same time. Then decide how many batteries you will be charging in each group and the interval between groups. (The interval is equal to 8 hours multiplied by the percent placed on charge. For example, if 5% are to be started at once, the interval is 5% of 8 hours, or 24 minutes.)

Figure 31 shows that when 5% are placed on charge every 24 minutes, the peak demand is only 59% of maximum. If your peak demand were (for example) 300 kW, the state-of-charge-based charging procedure would reduce the peak to about 177 kW. Table 6 shows the dollar impact of workshift-based vs state-of-charge-based charging.

 

Figure 30: Reduction of Peak Power Demand with Slate-of-Charge Charging Schedule

 

Table 6: Impact of Workshift-Based vs State-of-Charge-Based

Charging on Peak Demand Charges for a 20 Truck-Fleet

Charging Schedule

Power Demand

Number

of Batteries

at a time

Time between

Start of Charging

(Hours)

Peak

Demand

(kW)

Average Monthly

kW Demand

Charge (based on

$ 13.78*/kW)

20

10

5

2

1

8

4

2

. 8

.4

221

177

149

135

130

$3050.

2440.

2055.

1860 .

1790.

*$13.78 is the Summer kW Demand Charge levied by Consolidated Edison

which serves the New York Metropolitan area (7/80).

 

Capital and Labor Savings

Since each battery can last for at least one shift, and most last for part of a second, it is no longer necessary to have two batteries for every truck. The total of 40 batteries previously required for our fleet of 20 trucks is reduced to only 35, a 12% saving in capital investment. We reach this conclusion by evaluating the number of batteries required per shift as shown in Table 7.

 

Figure 31: Effect of Alternative Charging Schedules on Peak Power Demand

 

Table 7: Batteries Required per Shift

Class

Shifts

Per

Battery

Batteries

Per Shift

Number

of

Trucks

Spare

Batteries

Required

Next

Highest

Whole

Number

A

B

C

D

E

F

1

1.15

1.36

1.67

1.88

2.14

1

0.87

0.74

0.60

0.53

0.47

1

2

3

8

4

2

1 = 1

1.74 = 2

2.22 = 3

4.80 = 5

2.12 = 3

0.94 = 1

1

 

Spare Batteries .................. 15

Fleet Batteries . . . . . . . . . + 20

Total Batteries Required ..... 35

 

 

From the reduction in the required number of batteries there follows, naturally, a reduction in the number of chargers required, and in the number of square feet of space devoted to charging, changing, and maintaining of batteries.

There also follows from the reduction in the number of batteries and chargers a 30% reduction in the total number of battery charges required during the year's operation of the fleet.

Under the workshift-based procedure, 200 battery charges per week supported 20 trucks, 2 shifts per day. This amounted to 10,400 battery charges per year as shown in Table 8.

In the state-of-charge-based procedure, made possible by on-board monitoring of battery capacity, there are only 35 batteries (instead of 40) to support the fleet. Thus, in 52 weeks, only 6812 charges are required for the fleet, a net reduction of 3588 charges per year.

 

Table 8: Number of Charges Required Annually

to Support Fleet

Old Procedure

New Procedure

Class

Number

per Class

per Shift

Charges

per Week

Number

per Class

per Shift

Charges

per Class

per Shift

Charges

per

Week

A

B

C

D

E

F

________

Totals

1

2

3

8

4

2

________

20

10

20

30

80

40

20

________

200

x 52

10,400

1

2

3

8

4

2

_______

20

1

1.74

2.22

4.8

2.12

0.94

________

12.82

10

18

23

48

22

10

________

131

x 52

6872

 

Every time a battery is removed from a truck and charged a certain amount of labor is involved. In addition to the business of lifting, emplacing, and breaking and making connections, there is also the labor of checking specific gravity, topping off electrolyte level, cleaning, etc., all of which must be done every time a battery is charged, regardless of its state-of-charge when it is removed from the truck.

A net reduction of 3588 charges per year (more than 1.7 per workshift for our typical double-shift, 20-truck fleet) is certainly a labor saving worth examining on its own merits.

Further, since only a small number of trucks arrive at the charging station at any given time, owing to the fact that their batteries seldom reach full discharge at the same instant, there is a considerable reduction in waiting time as compared to the workshift-based procedure previously followed. Trucks (and drivers), therefore, are more productive on the average, since less time is spent standing idle, waiting for fresh batteries.