Operating Costs (based on a summary by R.S. Shevell)

The figure of merit used to evaluate competitive airplane designs is always based on a cost-benefit analysis. The minimum cost per unit of work performed must be the criterion, here the work is performed equally well by the competing designs. If one design excels in some aspect of its performance, i.e., goes faster, lands in a shorter field, makes less noise, provides more comfort, then a higher cost per unit of work may be justified. How much higher cost can be justified is always a difficult question and often involves a broader economic study of the system, e.g., costs of longer runways, or psychological and semi-economic judgments such as the value of a wider seat or greater speed. Even military aircraft can be judged on a cost effectiveness basis such as the total cost of delivering X troops to a location Y miles away.

The usual method of comparing the cost effectiveness of commercial aircraft is the direct operating cost, D.O.C.. Equations for estimating the comparative direct operating costs have been generated by the Air Transportation Association of America, ATA. First developed in 1944 from a paper published by Mentzer and Nourse of United Air Lines in 1940, these equations have been periodically revised in form and constants by the ATA to match current statistical cost data. The most recent issue was published in 1967 and is attached to these notes.

Direct operating costs can be expressed in terms of $/hour, $/mile, ¢/seat-mile, or for cargo aircraft, ¢/ton-mile. Costs in terms of $/mile indicate the maximum loss to an operator with an empty airplane, while costs per unit productivity such as ¢/seat-mile, or ¢/ton-mile are indicative of the fare that must be charged with reasonable load factors. Current practice usually bases costs on nautical miles although some people still like statute miles. It makes the D.O.C. look smaller. Figure 1 shows how aircraft DOC has changed with time in constant $ terms, illustrating the remarkable reduction in cost during the history of commercial flight.

Figure 1.

Direct operating costs are extremely useful for comparative analysis. Since the actual cost varies with accounting practice and with every change in fuel costs, labor contracts or parts price, obviously a perfectly precise D.O.C. method would have constants that changed with route, airline, and the time of day. The ATA method is based on the average of many airlines and can be expected to give a reasonable estimate of the average D.O.C. for the time period on which the statistical studies were based. Constants such as fuel cost per gallon and labor rate per hour can be adjusted for later periods or special circumstances. Regardless of the accuracy of the D.O.C. value, however, the equations can be expected to give a good comparison between different airplanes designed to the same state of the art. Figures 2 and 3 show how the price of gasoline and the average consumer price index have changed over the last many years as well.

Figure 2.

Figure 3.

The most intangible terms in the ATA equations are the maintenance quantities, specifically the labor man-hours per flight hour and per flight cycle for airframe and engines. The ATA values are based on experience but a new design can be better or worse than that experience. Because of the importance of these costs, a major engineering effort is applied to detail design to optimize accessibility, easy replacement, and selection of reliable components.

If a new design shows genuine improvement in maintenance characteristics, the maintenance costs may be estimated separately. Sound justification for the reduction in such costs from previous models — which presumably match the statistical equations — must be presented or the lowered costs estimates will have an impact limited to the consumption of ink on the pages of a 4-color brochure.

Examples of justified modification of the ATA equations by aircraft manufacturers are

1) Maximum engine parts replacement cost per hour is guaranteed by the

engine manufacturer. Then this value may be safely used in lieu of the equation.

2) The design uses significantly lower numbers of components than previous designs, e.g., fewer actuators, fewer valves, fewer switches — and the components selected have proven records of reliability.

Although aircraft manufacturers usually use the ATA equations, or slight modifications thereof, to make cost comparisons, airlines almost always generate their own equations based on each air1ine's individual cost experience. In spite of the differences that inevitably arise in these cost studies, the percentage each major cost item bears to the total cost is quite similar.

Figure 4 illustrates the relative importance of crew, maintenance, depreciation, insurance, and fuel costs by several methods for various airplanes. The method labeled "1966 ATA" was a preliminary and somewhat different form of the method issued in December, 1967. The data shown used engine manufacturer’s material guarantees. The 1967 ATA data shown have DC-10 and 747 maintenance costs reduced by approximately 20% due to design improvements. The European airline data use the airline’s own methods, the details of which are unknown. In spite of the diverse approaches, crew costs vary from 22% of the total at 100,000 lb take-off weight to 11% at 700,000 lb. (± ~1%), maintenance costs are about 27% of the total (± ~4%), depreciation varies from 22% for the small aircraft to 32% at 700,000 lb (± 2%), insurance is 6.5% (± 1%), and fuel is about 22% of the cost (± 2%) with the higher overseas fuel cost showing up in the European airline data.

Figure 4.

DATA SOURCE & METHOD

1. Based on 1966 ATA equations except engine manufacturers material guarantee

2. 1967 ATA cost method except DC-IC maintenance costs based on new Douglas estimates. 747 given equal % advantage over ATA

3. Major European airline, internally generated method

Figure 5 shows the primary parameters affecting D.O.C. as determined from analyzing the 1967 ATA equations. Relatively few variables are involved and it may be seen that many of these are interrelated. For example ‘no. total engines' is used to obtain either total thrust of all engines, or total engine cost -- which is itself, a rough function of total thrust. As shown in Figure 3, the cost per engine is a non-linear function of engine thrust so that 4 small engines cost more than 2 engines with the same total thrust. When aircraft being compared have the same number of engines this problem is eliminated.

PRIMARY AIRPLANE PARAMETERS AFFECTING DIEECT OPERATING COSTS

Primary Airplane Parameters

Thrust Cost Fuel
Block Take-off Airframe Total Airframe No. of Per Per Burned
Cost Item ($/mile) Speed Weight Weight Cost Cost Engines Engine Engine (lb/trip)

Crew Cost X X

Maintenance

Airframe X X X

Engine X X X X

Depreciation (1) X X X X

Insurance (2) X X

Fuel X

(1) Utilization and depreciation period also play an important role here.

(2) Utilization is important in insurance cost.

FIGURE 5.

Figure 6.

Airframe cost is directly dependent upon airframe weight, for equal ‘state of the art’ cases, and total cost is the sum of airframe cost and engine cost. Take-off weight is the sum of payload (assumed the same for all designs being compared) engine weight, airframe weight and fuel. Thus DOC. depends upon total thrust, airframe weight and fuel burned, and since these three items are the variables in take-off weight, a minimum take-off weight is a good ‘first’ guess criterion for minimum D.O.C.. Of course, the effect on D.O.C. of the interplay between smaller engines and larger airframe (wing area), or vice versa, to achieve the same mission at a given take-off weight cannot be seen without a detailed D.O.C. study. With a given set of engines, the least take-off weight will be very close to the minimum D.O.C., and even with variable engine configurations, a minimum take-off weight selection will usually come close to the most efficient configuration selection. For modern turbine powered transport aircraft, an estimate of airframe cost at $500 per pound of airframe weight is reasonable.

The exact impact of changes in individual airplane parameters on direct operating cost depends, of course, on the base case with which we start the analysis. The order of magnitude of the D.O.C. sensitivity factors is usually quite similar for different aircraft, however, so a good impression of the effect of design changes on D.O.C. can be obtained from the following table (Fig. 7) developed for the DC-l0-l0 trijet.

DIRECT OPERATING COST SENSITIVITY FACTORS

(Based on DC-10-10, 1967 ATA Method)

10%. INCREASE IN: % CHANGE IN DOC Airplane Total Cost (1) 5.0

Airframe Cost 3.8

Engine Cost 1.1 Airframe Weight 0.9

Block Fuel 2.0
Utilization -3.2
Flight Time 7.3
Insurance Rate 0.7
Depreciation Period -2.9
Engine Thrust 0.4
MAX. TOGW 0.2

(1) Both airframe and engine costs increased 10%.

Figure 7.

Figure 7 is based on pre-1973 fuel costs so that an updated set of sensitivity factors would show a higher % change in DOC due to 10% increase in fuel cost and a lower effect on DOC of other variables. In 1977 the fuel % change in DOC would be about 3.2 to 3.7 instead of 2.0.

Although direct operating cost as defined in the usual ATA derived method does not including landing fees, landing fees are an airplane related cost. Figure 5 shows a breakdown of direct operating costs in percentages for a proposed future airplane. The relative costs are shown for several fuel prices. The basic chart includes landing fees and shows them to be 11 to 12% of total direct costs. The figures in parentheses give the data for the usual direct costs without landing fees. For fuel prices ranging from l5¢/gal to 60¢/gal the percentage of DOC attributable to fuel varies from 23% to 54%.

FUEL PRICE

15¢/GAL 30¢/GAL 45¢/GAL 60¢/GAL

-------------------------------------------------------------------------------

Fuel 20% 33% 43% 50%

Depreciation 26% 22% 19% 17%

Insurance 7% 6% 5% 4%

Cockpit Crew 15% l2% l1% 9%
Maintenance 19% 16% 13% 12%

Landing Fees 13% 11% 9% 8%

----- ----- ----- -----

100% 100% 100% 100%

NOTE: BASED ON DC-1O TWIN 500 N Mi RANGE

Figure 8.

It is common to estimate aircraft manufacturing cost in terms of $ /lb. of airframe weight empty plus the engine cost. Actually each portion of the airframe has a different cost per pound. The table in Figure 9 shows the distribution of airplane costs between basic structure and the various aircraft systems as percentages of total airplane cost for a modern transport.

Basic Structure (Wing, Fuselage, Tail) 41.5%

Propulsion System including Engines 17.1%

Furnishings including Lighting 14.5%

Avionics (Communication and Navigation) 12.7%

Flight Control and Guidance Systems 5.3%

AC Power System 2.4%

Hydraulic and Auxiliary Power Systems 2.1%

Air Conditioning and Pressurization 1.9%

Landing Gear, Wheels, Tires, Brakes 1.7%

Miscellaneous Systems and Components 0.8%

Figure 9. Distribution of Airplane Manufacturing Costs

Direct Operating Costs are usually presented as curves of $/n. mile vs. range, and ¢/seat/mile vs. range, as shown in Fig. 10 for the DC-8 series 60 family. The break points in the ¢/seat-n. mile curves are the maximum ranges for which the full passenger capacity can be carried.

Figure 10.

Another example is shown in Fig. 11 where the $/stat-mile is plotted for very different sized aircraft. The DC-10 shown is the domestic version, the DC-10-10. The corresponding ¢/seat stat-mile is shown in the table of Fig. 12. Although the DC-l0-10 and B747 benefit in ¢/seat-mile from technology improvement, a significant part of the gain comes from larger size. This is emphasized in the $/mile data. In the comparison between the DC-10 and the B747, the B747 fails to gain in seat-mile cost even though it is much larger. This is partly because the improvement due to size flattens out between 300 and 400 passengers and partly because the B747 carries extra wing area and structure capable of much greater range than the DC-10-10. The overwater DC-l0-30 with a range comparable to the B747 would show about the same ¢/seat-mile as the B-747. Figures 13 and 14 show similar results based on more recent data.

Figure 11.

Direct Operating Cost Comparison

In Year 2000 $

Airplane (passengers) DOC (ct/seat n.mi) $/n.mi.

MD-81 at 500 n.mi. 6.15 8.79

MD-11 at 3000 n.mi. 5.81 17.03

747-400 at 3000 n.mi. 5.43 22.58

Figure 12.

Figure 13.

Figure 14.

INDIRECT OPERATING COSTS

Indirect operating costs (IOC) are those airline costs not directly connected with the actual flight of the aircraft. Indirect costs are just as real as other costs, but they are sometimes more difficult to separate and define. Indirect costs include the following:

Aircraft Ground Handling

Landing Fees

Aircraft Service

Cabin Attendants

Food and Beverage

Passenger Handling

Reservations and Sales

Baggage/Cargo handling

Passenger Commissions

Passenger Advertising

Cargo Commission

General and Administration

These costs are generally independent of the type of airplane, and thus are classified as indirect items. The actual value of each of these can only be estimated from statistics. A method of estimating the various factors has been developed by an Aircraft Industries Association Committee. Each term is based on maximum take-off weight, passenger capacity, enplaned passengers, or cargo carried, whichever is relevant to the particular term. For example the “Cabin Attendants” term is based on passenger capacity, while food and beverage, passenger handling, reservation, sales, commission and advertising are based on enplaned passengers. Since enplaned passengers enter into the equations, load factor influences the indirect costs.

Applying the IOC equations to the B-747, the DC-l0, and a typical large twin engine airplane show very similar relationships between TOC and DOC for all three airplanes. The average of these relationships is shown in Fig.15 in the form of:

TOTAL OPERATING COSTS (TOC) IOC

----------------------------------------------- = ----- + 1.0

DIRECT OPERATING COSTS (DOC) DOC

TOC/DOC is shown to be a strong function of range, load factor and whether flights

are domestic or international. TOC/DOC varies from about 1.7 for a domestic flight,

50% LF, and a range of 2200 nautical miles, to 2.1 for a domestic flight, 50% LF,

and a range of 400 nautical miles. International flight values with 50% LF, vary from 1.9 at 4500 nautical miles to 2.5 at 1100 nautical miles.

Figure 15.

The actual value of indirect costs may be estimated from an equation fitted to the results of the studies of the B-747, DC-l0 and the large twin mentioned above. The equation agrees perfectly with the detailed method at 50% load factor and shows only a 1 to 2% difference at 100% load factor. The equation gives the indirect cost in $/n. mile at a range of 1000 n. miles for domestic routes:

IOC_{1000 n.mi} ($/n.mile)= -.04 + .00129 W_g + .00119 N_p + .0l27 N_p LF (in 1968 $)

Where:

W_g = Maximum Take-off Weight (lb) / 1000

N_p = Passenger capacity

LF = Load factor

This equation is truly valid only for aircraft with cruise Mach numbers of about 0.85. However, speed differences of 10 to 20% will affect the IOC by only 2 to 4%. Higher block speeds reduce the IOC.

For other ranges the IOC is corrected by using the ratio of IOC($/n.mi) / IOC_1000nm ($/n.mi)

from Figure 16. The latter is derived from the same data as Fig.15.

Figure 16.

Breakeven Load Factor

To break even at distance, d , with a yield of $y /passenger-rnile, the revenue must equal the sum of the direct and indirect costs:

N* LF * d *y = DOC * N * d + N * LF * ($/pass)_indirect

where DOC and ($/pass)_indirect are taken at distance, d ; N = number of pass. seats. LF is the breakeven load factor. Substituting:

LF_breakeven = d DOC / [y d – ($/pass)_indirect]

Total operating costs may be used in a complete airline system analysis in which each city-pair is studied to determine total traffic, required schedule frequency, load factors, total income, total costs and profit. Simpler presentations of the effect of costs may be shown in the form of passenger load required to pay the DOC as shown for the B707-320B and the B747 in Figure 17. Another type of analysis determines the break-even load factor, the load factor required to cover the total costs. Figure 18 shows this type of analysis for the DC-10, B747, DC-8-62, and the B727-200. All three of these economic analyses require establishing not only operating costs but also the yield, the average passenger fare per mile. The yield varies greatly with route and is generally different from the basic fare as airlines now determine fares based on the day of the week, when the ticket is purchased or whether the traveler will stray over a Saturday night.

Figure 17.

Figure 18.