3. Methodology

Since records on national and provincial level related to commercial harvesting in Mozambique were scant, data to analyse efficiency had to be collected on-site. Therefore the case study was chosen as the methodological framework for this investigation. In order to take into account distinctive bio-physical and commercial conditions, five companies were selected in the northern (Cabo Delgado, Nampula), central (Sofala, Zamb�zia), and southern (Gaza) provinces of Mozambique. Criteria for selection were that the enterprise should be of small scale and harvesting timber throughout the period of the field studies. Data were collected by the author during five field visits of one week at each enterprise, in November 1996, October 1997, and July 1998.

According to SILVERSIDES & SUNDBERG (1989) efficiency in forestry is defined as the effective use and economical management of forest resources. The same authors emphasise that efficiency describes the quality of production, indicated by the ratio between two parameters (e.g., cubic metres of extracted timber per unit of extraction time). The present study considers efficiency to be an indicator of the viability (and hence sustainability) of timber production.

Figure 3.1 reflects the methodological approach used in this study. Data collection relied on work studies and the appraisal of equipment and personnel. Together with a descriptive analysis of the processing sequence within work cycles, the collected data were put into ratios, resulting in operating time per cubic metre, including mean production volume and distance, and in costs per machine-hour. On the basis of these preliminary results efficiency was analysed. To this purpose a set of indicators was conceived that relate production data (e.g., log volume) to consumption data (e.g., time), or, as L�FFLER (1990) suggests, that establish a ratio between output and input. Operational, organisational, energy, and financial output data were rated with input data of the same origin, quantifying the extent to which the indicators measure efficiency.

3.1 Collecting operational and financial data

Field studies concentrated on collecting operational and financial data that would be essential for subsequently quantifying efficiency indicators. For this purpose work studies in logging and transport operations were conducted. Additionally, directors and managers were interviewed in order to obtain basic data on equipment and personnel.

3.1.1�Time and production study

Time and production data were recorded by means of time studies. Time studies proceed by structuring a specific activity into repetitive cycles. Every cycle is made up by elements. A time-study record registers the type and measures the duration of elements occurring in work cycles. By means of recorded time data, the mean duration of a cycle is calculated and rated with the mean volume produced in the cycle, yielding operating time per production unit and technological productivity in the studied activity (KANAWATY 1992). Table 3.1 illustrates how the activities observed during the field studies were structured into repetitive work cycles and defined, codified elements.

Time studies not only measure time and production, but also identify time types according to the elements occurring. Total time recorded is subdivided into main time (productive) and general time (unproductive) (REFA 1991). Main time appearing in productive processes is separated into effective time that occurs during production (e.g., travel with load), and auxiliary times (e.g., fasten chain to log in skidding). General time that interrupts the productive process is divided into times for preparation and conclusion, maintenance, rest and technical/personal interruptions. L�FFLER (1990) refers to these times as �non-cyclic times�. KANAWATY (1992) uses the term �inefficient time�, and APUD & VALDEZ (1995) speak of �secondary time� (as distinct from �principal time�). The present study prefers the notion of �unproductive time� (as distinct from �productive time�). Note that these terms are synonymous and specify those times occurring in elements which cannot be associated to one cycle, and are therefore expressed as a percentage of total time of all recorded cycles.

�(a) Preparing time and production records

In a preparatory phase the elements constituting a work cycle as well as their initial and final timing points were defined. Furthermore, the beginning and end of the repetitive cycles that jointly constitute the activity had to be determined. Each element was specified by a code, a description with its associated time type, and the initial and final point of timing. Table 3.1 shows the specifications as defined in this study for the analysed activities.

From the definitions in Table 3.1 it was possible to attribute a time to each component element of a specific activity, to evaluate its duration and portion of total time, and to distinguish productive from unproductive times. In all five companies the same element codes and time-type associations were used for the time studies in order to facilitate cross-company comparisons.

(b) Conducting the time studies

The time studies were designed to record both element times and their chronological sequence in order to obtain a detailed image of how production was performed. For this purpose an electronic chronometer was used in continuous timing (CT) as described by KANAWATY (1992). In comparison with other timing methods, continuous timing has the advantage of not only registering separate element times but also facilitating analytic reconstruction of the sequence in which elements occur during the operations (REFA 1991).

An electronic chronometer RUCANOR style='font-family: Symbol'>� was used, which has a digital display that simultaneously measures both partial times and accumulated time in minutes and seconds. Time study was started by activating the chronometer. At each timing point (end of a work element) the partial time (time elapsed from the previous timing point) was displayed by pressing the lap-time button. The element time and the code of the elapsed element were recorded (see 3.1.1(c) below). Meanwhile the accumulated time continued to be measured internally and appeared again on the display whenever the split-time function was disabled. At the end of time study the last partial time was recorded and the chronometer was stopped, indicating total elapsed time for the entire timing period.

The beginning and end of the time study were recorded (in daytime hours and minutes) with a normal watch. Subtraction of the beginning daytime from the ending daytime yielded a control time (REFA 1991), which was then compared to the total elapsed time during the study. As recommended by L�FFLER (1990), time-study records with a divergence between total accumulated time and control time of more than 3% were rejected.

Reference data for subsequent calculation of efficiency indicators were measured and recorded in each observed cycle. Depending on the activity, distances between adjacent felled trees, species, DBH, tree height, stump height, log lengths, log diameters at centre points, and extraction and transport distances were recorded. A reversible metric tape and a diameter tape were used for measuring log lengths and diameters.

�(c) Recording data

A form was developed for recording the sequence and duration of work elements as well as reference data. In the first column the cycle number was registered, in the second column the element code, in the third column the partial time in minutes and seconds, and in the fourth column the partial time in centiminutes. The remaining columns were dedicated to observations and reference data.

The header included cells for page number, production unit and team, observed activity, calendar date and control time. After having filled all 35 lines of a page, accumulated time was registered in the footer. The remaining space on the last page served to calculate total and mean values of time and production in recorded cycles.

3.1.2�Appraisal of equipment and personnel

Directors and managers were interviewed in order to obtain data on equipment and personnel employed in production. For this purpose a form was developed to identify equipment types and purchase prices as well as the number of workers and their annual wages in logging and transport (including roading), maintenance, processing, and supervision.

For subsequent cross-comparison of the companies� financial efficiencies, baseline data for the following parameters were determined. Table 3.2 compiles the results in summary.

�(a) Annual working time and wages

Annual production period varied, depending largely on local climatic conditions, between eight and nine months. Cost calculations were based on the assumption that companies produced eight months a year and that during the rainy season the logging area was not accessible. In general, 20 working days per month and nine hours of daily working time yielded an annual working time (AWT) of 160 days or 1440 hours. Since rains had little affect on AWT for maintenance crews, a production period of nine months was determined for these crews, equivalent to 1620 hours. SOMANOL�s sawmill worked eleven months a year and nine hours a day, resulting in an AWT of 220 days or 1980 hours.

Data on wages reflect the system applied in each company. Total crew wage per hour was calculated on the basis of the number of crew members and their annual wages in Meticais (Mt), divided by total annual work time and converted into US dollars (US$) per hour.

�(b) Depreciable value

Purchase prices were recorded as indicated by each company�s management. In order to obtain depreciable values the price for consumable commodities (e.g., tyres) and the residual value, generally assumed to be 10% of the purchase price, were subtracted. Where very old equipment was used it was assumed not to have any residual value at the end of depreciation period.

Table 3.2 Baseline data for efficiency analyses.

Treating the repair rate as a percentage of depreciation allows evaluating repair costs during the machine�s useful life. According to recommendations by GRAMMEL (1988), repair factors of 1.0 were used for extraction and transport vehicles, 0.8 for chainsaws, and 0.5 for sawmill machinery as well as equipment used for loading, road maintenance, and workshop, and for vehicles used in supervision. These assumptions seem to reflect real production patterns, since the equipment had been purchased second-hand, having already been used extensively and exposed to heavy wear by abrasive sand and corrosion.

�(c) Costs of consumable commodities (fuel, lubricants and tyres)

Fuel prices varied among provinces and during the years in which data were collected. In order to keep companies comparable to one another, a standard price was determined and applied to all calculations. From 1996 to 1998 costs averaged 5,500 Mt (0.46 US$) per litre of diesel, 6,500 Mt (0.54 US$) per litre of gasoline and 36,000 Mt (3.00 US$) per litre of lubricant.

Fuel consumption (L_h) was estimated on the basis of company records when they were available. Otherwise it was calculated as a function of engine power (P_hp) as suggested by FAO (1992):

(1)
������������ ������������

where���L_hp� = fuel consumption rate per horsepower = 0.17 kg/hp�h for diesel and 0.21 kg/h�ph for gasoline
������������P_hp� = engine power, hp
������������F_c��� = load factor, estimated to be 0.54 for tractors and chainsaws and 0.38 for trucks and generators.
������������W��� = fuel weight = 0.84 kg/l for diesel, 0.72 kg/l for gasoline

It was assumed that machine engines consume lubricants equivalent to 10% of fuel consumption. Costs for tractor, truck, and trailer tyres were determined according to the companies� records.

�(d) Rates of interest and exchange

From 1996 to 1998 effective interest rates varied depending on financing sources. To facilitate comparison among the companies, a uniform effective interest rate of 15% a year was applied. Due to particular funding conditions in the case of ECOSEMA, costs for that company were calculated with an interest rate of 25%. From 1996 to 1998 exchange rates for the Meticais ranged between 11,500 Mt and 12,500 Mt per US$. Cost calculations applied a uniform exchange rate of 12,000 Mt for 1 US$.

3.2 Data processing � efficiency analysis

Recorded data were processed in order to calculate efficiency ratios. This section identifies the indicators that were used for efficiency and describes how they were evaluated and quantified on the basis of recorded data, comparing outputs against inputs. Distinctions are made among operational indicators derived from time and production data, organizational indicators evaluating synchronization and coordination of equipment and workforce, and financial indicators based on the calculation of unit costs. Additional indicators were used to estimate energy efficiency.

Compilation and processing of data collected in the time studies was carried out with Microsoft� Excel. Machine rates (costs per hour) were calculated using PACE (FAO 1992).

3.2.1�Operational efficiency

Operational efficiency is defined by the equilibrium of procedures and equipment in all production phases, in order to maintain production on a sufficient level and to avoid interruptions (SILVERSIDES & SUNDBERG 1989).

The present study evaluates operational efficiency in logging and transport. Due to the limited time available for data collection, processing of logs into final products was not analysed.

In total 256 cycles were observed. Table 3.3 summarises the number of cycles, broken down into activities by company, and shows the statistical variation in the data as collected. Transport occurred with low frequency during data collection. Therefore only nine cycles were observed for first transport and seven cycles for second transport.

Table 3.3 Number of cycles (coefficient of variation in %) for the analysed activities.

Operational efficiency was evaluated by processing relevant time and production data collected during the time studies. The sequence of elements as well as their temporal distribution within the work cycles was identified. Technological productivity was calculated from operating time and production volume. And finally, the ratio between tree volume and log volume yielded the recovery rate associated with primary production.

�(a) Element sequence and temporal distribution

The order in which element codes were listed in the record form reflected the sequence of work elements within observed activities. In order to evaluate the temporal distribution of elements, both total productive and unproductive times and their shares of total recorded time were calculated. Then the portions of total productive and unproductive time occupied by each element were calculated. In addition, cycle elements that consumed time disproportionately, and thus constituted critical elements, were identified for each operation.

�(b) Technological productivity and recovery rate

Recorded total time (T_t) (without interruptions longer than 15 minutes) and mean log volume (V_m) per cycle were calculated and put into relation, yielding operating time (T_o) expressed in minutes per cubic metre. Technological productivity (PT) was obtained by dividing the number of minutes in one hour by operating time, expressed in cubic metres per machine-hour (L�FFLER 1991):

(2)
������������ �����������������or
������������ ������������

PT is one of the most frequently used measures of operational efficiency in a specific activity. However, it only reflects procedures during recorded production periods. Non-productive periods, which under prevalent conditions occupied a large portion of annual working time, are not measured by this indicator (see Section 3.2.2(a)).

Additional reference data were processed to determine logging intensity in cubic metres per hectare as well as average distance and average speed in extraction and transport.

Recovery rate was evaluated comparing log volume and commercial tree volume, based on the assumption that commercial tree volume contains wood for commercial purposes to a diameter of 20 cm. The tree volume (V_t) was estimated by multiplying basal area, total tree height (H_t) and a taper factor (f_t) that varied between 0.4 and 0.6 depending on the particular species. Thus commercial tree volume was calculated as follows:

(3)
������������ ������������

where p = 3.14159�, the ratio of the circumference of a circle to its diameter. Similarly, log volume (V_L) was calculated by the formula:

(4)
������������ ������������

where D_m�is the log diameter at the midpoint of the log, and H_L�is the log length. Then the ratio:

(5)
������������ ������������

yields the average recovery rate (RR) expressed in percent.

3.2.2�Organisational efficiency

This set of indicators is related to synchronization and coordination of activities that are essential for maintaining production on a sufficient level, and expresses the relative efficiency with which machinery, workforce, and capital are employed.

�(a) Utilisation rate and production capacity

According to L�FFLER (1991), utilisation rate is intended to measure the efficiency with which a company uses its machinery, indicating the quality of synchronisation among activities and machinery. Since company records on annual machinery utilisation were not available, the present study estimated the production capacities of equipment involved in logging and transport (in cubic metres per year) by calculating the product between number of machines used in the specific activity (N_e), technological productivity per machine (PT), and total annual work time (TT_a). Utilisation rate (UR) was then obtained by comparing annual production volume (E_a) with technological capacity, through the formula:

(6)
������������ ������������

UR, expressed in percent, indicates the average efficiency of applying equipment in a specific activity during one year. With regard to the entire production, the equipment with the lowest production capacity determined the company�s annual production capacity and served as reference for overall utilisation rate.

�(b) Labour productivity

To estimate workforce productivity (P_L), the timber volume (E_a) produced in the year in which the company was analysed was divided by the number of the workers (N_p) in logging, extraction, and transport, and the number of months (N_m) worked during that year:

(7)
������������ ������������

The result, expressed in cubic metres per worker-month, is a measure of the efficiency with which the workforce was employed in production.

�(c) Capital intensity

Capital intensity (IC) indicates the amount of money with which the company�s investments debit each cubic metre of timber produced. Although linked with financial parameters, IC is an organisational indicator showing the extent to which the company was able either to minimize the sum of mean annual investments (I_ma) or to maximize annual production (E_a). Capital intensity, expressed in US$ per cubic metre, was calculated with the formula:

(8)
������������ ������������

3.2.3�Energy efficiency

In addition to labour, land, and capital, it is becoming more and more common to include energy among the most important production factors in forestry. The concept of energy balance (L�FFLER 1991) covers consumption of direct energy, human energy, and energy spent in machine manufacturing. The present study evaluates energy efficiency on the basis of direct energy input by fuel consumption in logging, transport and processing. On the output side, the potential energy represented by timber products is considered.

First, energy consumption per hour (L_h, see 3.1.2(d)) was converted to energy consumption per cubic metre of timber produced within the specific activity (L_u = L_h�/ PT). Then the specific consumption values for all machines involved in production were summed for logging and first transport and for subsequent activities (second transport and processing) and expressed per 1,000 litres of fuel, using the formula:

(9)
������������ ������������

where style='font-family: Symbol'>h _e�is expressed in cubic metres of timber produced per 1000 litres of fuel.

According to L�FFLER (1991), one cubic metre of timber contains energy corresponding to a calorific value of 200 kWh, and the energy content of one litre of fuel is around 10 kWh (diesel 9.88 kWh/l, gasoline 9.72 kWh/l). Thus, an energy coefficient (h_c) can be calculated with the formula:

(10)
������������ ������������

Both Equations (9) and (10) show the ratio between energy that results from products (logs, sawn timber) and the direct energy that is consumed as fuel during production, providing an indication of the efficiency with which fuel energy was employed. In this context, designing an energetically efficient process means using machines with low fuel consumption and maximising technological productivity, particularly in transport and processing.

3.2.4�Financial efficiency

Indicators frequently used to evaluate the viability of an enterprise include production costs and their relation to sales revenues. Unless sales revenues cover the cost of production, a firm�s activities are deficient and economically unable to sustain production, to create value added, and to contribute to the national economy (PRETZSCH 1997).

The present study is based on a model that evaluates harvesting and processing costs, without considering royalties or additional costs eventually associated with silvicultural activities. Subtracting costs from sales revenues yields profit margin, i.e., the monetary value that the resource renders.

The results of this analysis indicate the degree to which production generates profits and is economically efficient, and the production level at which costs would be covered by sales revenues.

�(a) Costs per machine-hour

By means of the computer program �Production and Cost Evaluation � PACE� (FAO 1992), data on equipment and personnel (compiled in a standardised form, see Section 3.1.2 and Table 3.2) were processed, resulting in costs per machine-hour for primary (logging and first transport) and secondary activities. Digital data processing for each activity resulted in a compilation of ownership costs, operating costs, and labour costs, as well as an overall summary. The types of costs and formulae used by the PACE program are compiled and described in Table 3.4.

Table 3.4 Cost components and their calculation.

The machine-hour (see Section 3.2.1), as a reference unit for calculating machine costs per hour, is derived from annual utilisation. A machine-hour corresponds to one hour�s use of a particular machine, including interruptions of less than 15 minutes (L�FFLER 1991). Conversely, wages refer to total annual working time. In order to be compatible with ownership and operational costs, PACE converts them into labour costs per machine-hour (see Equation (23)), thus avoiding fragmentation of the results, referring the costs for both equipment and personnel to the machine-hour. Furthermore this conversion creates a uniform reference for calculating unit costs for all activities, dividing the costs per machine-hour by technological productivity (see Section 3.2.4).

Additional processing of financial data was carried out with Microsoft � Excel. Costs per machine-hour for primary activities (logging and first transport) as well as secondary activities (second transport, maintenance, and processing) and supervision were compiled. Whenever two or more equipment/personnel units were used for the same activity, respective costs were added accordingly; e.g., in truck loading with tractor assistance, the operational costs of the truck were not integrated into the cost scheme.

�(b) Comparative unit costs (logging and first transport)

Technological productivity evaluated on the basis of time-study data (Section 3.2.1(b)) were used to derive unit costs. Dividing the cost per machine-hour for a particular activity by its technological productivity yielded the activity�s unit cost in US$ per cubic metre. Then ownership, operating, labour, and total costs in US$ per cubic metre were compiled for logging and first transport. Subsequent activities were not included, as they did not occur in all cases (second transport, processing) or were not directly associated with production (maintenance, supervision).

The resulting unit costs provide an indication, under uniform conditions, of the relative efficiency of equipment and personnel when producing one cubic metre of logs. The influence of annual machine utilisation on unit costs was neutralised by assuming that in all cases. equipment and personnel were employed according to the standardised annual utilisation (e.g., 800 machine hours in skidding, which corresponds to an utilisation rate of 56%). Thus companies� activities and their cost structures could thus be compared with one another, independently of their actual production level. However, comparative unit costs assume conditions related to machine utilisation that do not necessarily reflect the company�s actual production pattern.

�(c) Effective unit costs and margin of profit

Effective unit costs are those associated with an entire year�s production, expressed in US$ per cubic metre of timber produced. Fixed and semi-variable costs were calculated by dividing annual ownership costs (C_fa), annual repair costs (C_ra�= C_r�� U_a), and annual labour costs (C_la = C_lh�� U_a) by the volume produced (E_a). Purely variable costs (fuel, lubricants, tyres) were added, dividing variable costs per hour without repair costs (C_vh�- C_r) by technological productivity (PT). Summing these yields the effective unit cost of production (C_au) for a specific activity:

(24)
������������ ������������

Activity costs were then summed to derive total effective unit costs (C_tu). In order to facilitate comparison among the companies, these total effective unit costs were subdivided into those associated with primary activities (logging and first transport) and those incurred in subsequent activities or in support (second loading and transport, maintenance, non-operational equipment, processing and supervision).

Sales revenues from final products (R_s), obtained from company records, were then compared to the total effective unit costs in order to evaluate the margin of profit (M_p) per cubic metre of timber produced:

(25)
������������ ������������

where M_p is expressed in US$ per cubic metre.

Finally, effective unit costs were interpreted as to their origins. As formula (24) suggests, there are basically two components determining the magnitude of effective unit costs: the ratios between annual fixed costs and production volume, and between fuel consumption per machine-hour and productivity. Consequently, effective unit costs provide a measure of the efficiency with which the company was able either to minimise heavy-machine use and fuel consumption or to maximise annual production and technological productivity.

�(d) Break-even point

The break-even point marks the volume of production at which effective unit costs are covered by sales revenue. Its level depends on the magnitude of annual fixed costs (ownership, repair, and labour) and the ratio between fuel consumption and technological productivity.

In order to analyse unit costs as a function of the volume of timber produced, the costs of ownership, repair, and labour (C_f, influenced by production volume) were multiplied by annual production volume (E_a) and then divided by the potential production volume (E_p). Then purely variable costs (C_v�= C_vh�- C_r) were added. Thus the unit costs of a proposed production pattern (C_pu, in US$/m³) were:

(26)
������������ ������������

Following OSWALD et al. (1997), the break-even point (BEP) where production costs are covered by sales revenues, was calculated with the formula:

(27)
������������ ������������

where C_fa = annual ownership costs
������������C_ra = annual repair costs
������������C_la = annual costs for wages
������������R_s� = sales revenue per cubic metre
������������C_v = variable costs excluding repair costs (C_vh�� C_r)

Finally, the break-even point was compared to the company�s technological capacity, in order to determine whether the break-even volume could have been produced with the existing equipment.