Time of day impacts on machine productivity and value recovery in an off-forest central processing yard
© Murphy et al.; licensee Springer 2014
Received: 19 November 2013
Accepted: 31 July 2014
Published: 5 September 2014
Effective use of the high capital cost equipment in a central processing yard requires a good understanding of the human component of the system. Poorly designed work schedules have been linked with mental and physical fatigue of the machine operators, low productivity and low value recovery for some operations. The aim of this study was to determine if time of day impacts machine productivity and value recovery in an off-forest central processing yard.
A database, containing over 120,000 records on Pinus radiata D.Don (radiata pine) stems processed during 214 work shifts, was analysed in order to determine the impact of time of day on value recovery and productivity of log-handling equipment that consisted of a scanning optimiser and two mechanised processors operating in an off-forest central processing yard in New Zealand.
Analyses indicate time of day negatively impacted volume productivity and value recovery for the scanning optimizer between the first shift operating mainly in daylight hours and the second shift operating mainly during night hours. There were no clear trends in productivity for the mechanised processors.
These findings are in agreement with an earlier study carried out in on-forest mechanised harvesting operations in Chile but differ from findings of another off-forest central processing yard in New Zealand.
Traditionally, initial assessment of logs in New Zealand has occurred at landing areas in the forest close to felling operations. Another approach used intermittently in New Zealand for over forty years has been to transport logs to a central processing yard (CPY), sometimes called a log-sort yard. Purported advantages of centralised processing include the provision of many wood-marketing and value-capturing services via concentrating, merchandising, processing, and sorting of logs (Dramm et al. ). Additional advantages may include savings in road construction and transportation costs, reduced logging and landing construction costs, reduced environmental impacts resulting from smaller landings, improved utilisation of logging residues, and greater opportunities resulting from artificial lighting to work longer hours in a yard than in a forest (McKerchar & Twaddle ).
Extended working schedules (i.e. those longer than a standard eight hours) are commonly used in the sawmilling and pulping sectors of the forest industry globally but have had mixed success in the harvesting, stem conversion, and transportation sectors. They have been tried and discarded in some parts of the world but have been used successfully for many years in other parts. Extended work schedules may include working more hours per shift, multiple shifts per day, more days per week, or some combination of these. Increasing production capacity, production efficiency and monetary returns are reasons given for operating extended hours.
Depending on their design, CPYs can require high capital outlays. Effective use of the high capital cost equipment in a CPY (or in forest operations) requires a good understanding of the human component of the system. Kirk () noted that studies worldwide have linked poorly designed work schedules with mental and physical fatigue, low productivity and low value recovery. Many researchers from a wide range of industries, including forestry, have found that hourly production declines as shift length increases (Vernon , Golsse , Nevison , Hanna et al. , Passicot & Murphy ). Productivity can also be lower for night shifts than for day shifts. Kerin and Carbone () reported an average drop in productivity of 5% for night shifts across all major U.S. industries based on surveys of employees and managers from over 1000 companies. Studies of forest harvesting operations in North America and Australasia reported declines in hourly productivity as high as 40% for night shifts compared with day shifts (Maxwell , Nicholls et al. , Mitchell ).
Murphy and Vanderberg () noted that, while there was potential for a reduction in costs resulting from increased daily production by working extended hours, the size of the production increase was sometimes insufficient for cost reductions to be realised. The impacts of extended hours on other tangible and intangible costs such as value recovery losses and human factors (e.g. employee-turnover rates, accident risk, and the opportunity for employees to participate in social and domestic activities) need to be considered (Mitchell ).
Research on hundreds of World War I munitions factory workers in the United Kingdom by Vernon () allowed him to examine the effects on productivity of a wide range of factors, including age, sex, season of the year, length of shift, and time of day. As a result of his research, he recommended studying as large a group of workers as possible over as long a period as possible when investigating output from alternative work schedules in order to remove physical and psychological factors for individual workers,. He also commented on the use of indirect observations (e.g. taking measurements from machines being used by the workers) to supplement direct observations. These recommendations were made because: (1) there is wide variability in the time of day at which humans perform at their best; and (2) the action of being observed may positively or negatively influence the performance of the observee ("observer" effect) for short-duration studies.
This paper reports the results of a case study on the effects of extended working hours on the productivity and value recovery of an off-forest, central processing yard in New Zealand. Results are based on long-term data that have been collected by indirect methods.
Pan Pac Forest Products Limited (Pan Pac) is an integrated forest products company that owns 33,000 hectares of forest plantation on the east coast of the North Island of New Zealand. Their Forestry and Logistics Division manages an annual volume of 1.5 million m3 of which about 0.75 to 0.90 million m3 comes from their own estate. Pan Pac produces pulp, timber, export chip and export logs from its operations.
This study included the scanner operators and the log-bucking processors but excluded the loader operators. The Logmeister system is in operation five days per week with two shifts per day. The first shift for the scanner operator is a fixed length of 9.5 hours and runs from 4.00 am to 1.30 pm. Minor maintenance and refuelling can be done between 1.30 pm and 2.00 pm by operators from the first shift. The second shift is open-ended. It runs from 2.00 pm until the time when all the stems produced from the forest estate that day have been scanned. This is often around 10.00 pm but can extend to 3.00 am the following day. A single scanner operator is scheduled for each shift, but scanner operators sometimes swap roles with other machine operators or manual workers. One, sometimes two, processors are scheduled to cut the stems into logs per shift. Shift lengths for the processors vary from 8 to 12 hours. One 30-minute rest- and meal-break per shift is taken by both scanner and processor operators. Competition exists among workers between the first and second shifts about who can produce the most wood volume per shift. Although not company policy, there was a tendency during this study for the loader operators from both shifts to feed bigger stems into the system at the beginning of the shift to quickly build up volume, and the smaller stems were left over towards the end of the shift.
Data records from early January 2011 through to the end of May 2011 were extracted from the Logmeister database for analysis. During this period, 120,807 radiata-pine stem pieces with a total volume of 246,844 m3 were scanned and processed over 200 work shifts. Piece lengths ranged from 1.27 to 20.11 m (mean 12.60 m). Piece volumes ranged from 0.04 to 11.03 m3 (mean 2.04 m3).
Number of pieces scanned or processed per machine operator
Number of pieces
Proportion of total (%)
Number of pieces
Proportion of total (%)
Six people operated the processors cutting the pieces into logs during the study period. One operator was excluded because of the small number of pieces he processed. The remaining five operators processed 99.3% of the pieces (Table 1). All processor operators were men. Experience on the processors ranged from 1 to 4 years.
Scanner operators and processor operators were all allocated to rotating shift schedules; that is, all scanner operators and all processor operators worked in both the first and second shift at some point during the study period.
Although there were two shifts for machine operators, managers and company and contractor supervisors worked just the traditional "9 am to 5 pm" day. Production and value performance was scrutinised more intensively during the hours that managers and supervisors worked.
Production data from the Logmeister database were imported into a Microsoft Excel™ spreadsheet. Production data included piece count, piece lengths, volume, quality codes called by the scanner operator, time to scan, and time to process pieces. The database did not include reporting of major delays, nor their causes. Pivot tables were then used to summarise the production data by time of day (hourly intervals), operation (scanning or processing), operator identity (ID), and first shift/second shift classification. Multiple regression analysis, using indicator variables, was used to examine the effect of key parameters (e.g. operator ID, time of day, average stem size) on machine productivity (m3 h−1). A parameter was considered to be significant if the p-value was less than 0.05.
Log specifications and prices used in the value recovery analyses
Relative Price ($ m−3)
Maximum branch diameter (cm)
Minimum small end diameter (cm)
The total number of stems scanned during the first 9.5-hour shift (47%) was similar to the number scanned during the first 9.5 hours (2.00 – 11.30 pm) of the second shift (46%). The remaining 7% of the stems were scanned in the five hours between 11.30 pm and 4 am the following morning. These late evening/early morning hours were worked only when not doing so would have resulted in an "insufficient" stockpile of scanned stems to meet processor and mill needs for the beginning of the following day.
Regression model for scanning piece count productivity (pieces h −1 )
Group A operators
Group B operators
Total degrees of freedom
Adjusted R Square
Regression model for scanning volume productivity (m 3 h −1 )
Group A operators
Group B operators
Time of day3
Total degrees of freedom
Adjusted R Square
Regression model for processor volume productivity (m 3 h −1 )
Group C operators
Group D operators
Total degrees of freedom
Adjusted R Square
Regression model for scanning value recovery ($ m −3 )
Stem Size Squared
Group E Operators
Time of Day3
Total degrees of freedom
Adjusted R Square
Discussion and conclusions
No consistent impact of time of day on overall productivity was evident from the data studied. Scanner volume productivity was found to be negatively correlated to the time of day. However, statistically significant impacts on scanner piece count productivity or processor volume productivity due to time of day could not be discerned. Instead, differences were due to average piece size being handled, number of machines being used, or differences between operators.
Rose et al. () found no drop in productivity for the night shift compared with the day shift of a large, non-mobile, centralised processing yard in New Zealand. Very good lighting outside of normal daylight hours was a feature in both the yard studied by Rose et al. () and the current study. The current findings for the processor volume productivity and the scanner piece count productivity are in agreement with those findings of Rose et al. () but not for scanner volume productivity.
The scanner volume productivity results, but not the processor volume productivity results or the scanner piece count results, from the current study of an off-forest central processing yard in New Zealand agree, however, with the findings of a recent study of on-forest mechanised harvesting operations in Chile. The Chilean study showed that productivity (m3 h−1) was negatively affected by working extended hours and/or multiple shifts per day (Passicot & Murphy ). As noted in the Introduction, other researchers have also found shift length and time of day impacts on on-forest operations productivity (Maxwell , Golsse , Nicholls et al. ).
The current results indicated both volume productivity and piece count productivity were significantly higher for the hours when two or more operators shared the scanning task than for those hours when a single operator undertook the scanning task. Gellerstedt () noted that Swedish experience has shown that high levels of productivity can be sustained throughout the day by rotating jobs within forest-harvesting crews and by allowing operators to select the day or evening shift that suits them best in a multi-shift operation.
Comment is often made in the literature on the effect of circadian rhythm on error rates which are at their highest between midnight and 6 am, peaking in the early hours of the morning (2.00 to 4.00 am) (Folkard and Tucker , de Mello et al. ). It was expected that increased error rates would lead to lower value recovery during this period of the day. In the current case study, value recovery was found to decrease at the rate of about 0.3% per hour throughout the day between midnight and midnight. The present authors are unable to provide reasons as to why value should continue to drop at the same rate between 6.00 am and midnight. It is worth noting that scanner operators had good overhead lighting and operating conditions, and the use of scanning optimising software, not the human scanner operator, determined the log-cutting regime. The operator(s) did need to identify and mark changes in quality along the stem but they did not have to decide what log types should be cut. Future research should explore the impact of time of day on value recovery for processors operating on-forest, particularly if these are not fitted with an optimising computer.
Statistically significant differences in value recovery, albeit small (~5%), were noted among scanning operators. The range in value recovery among operators is considerably smaller than that reported by Murphy () based on worldwide studies of mechanised operations. This may be due to good operator selection on the part of the Logmeister contractor or due to the use of an optimiser on the scanner requiring fewer log-making decisions by the operator.
As noted in the Introduction, Vernon () recommended using indirect observations gathered over as long a period as possible to study as many people as possible. In this study, indirect observations gathered over a five-month period were used to study five scanner operators and five processor operators. While sample sizes of five are better than one, the operators included in this study may not reflect the true range of forestry machine operator reactions to different work schedule designs. The authors also note, however, that the Logmeister system was the only one of its type operating in New Zealand at the time this study was undertaken and all operators working more than a few hours were included in the study.
Further work is needed on time of day impacts and work schedule design on production economics. Understanding the effects of extended work hours and different work schedules on people, productivity and value recovery of both on- and off-forest mechanised operations will allow planners to better manage log supply, labour force requirements, and the economics of the forest to mill supply chain.
AD participated in the collection and interpretation of the data and assisted with the drafting of the manuscript. HM participated in the preliminary analysis of the data and assisted with the drafting of the manuscript. GM participated in further analysis and interpretation of the data and drafted and revised the manuscript. All authors read and approved the final manuscript.
- de Mello MT, Esteves AM, Pires MLN, Santos DC, Bittencourt LRA, Silva RS, Sufik S: Relationship between Brazilian air pilot errors and time of day. Brazilian Journal of Medical and Biological Research 2008, 41: 1129–1131. 10.1590/S0100-879X2008001200014View ArticlePubMedGoogle Scholar
- Dramm JR, Govett R, Bilek T, Jackson GL: Log sort yard economics, planning, and feasibility (General Technical Report FPL-GTR-146).. Department of Agriculture, Forest Service, Forest Products Laboratory, Madison, WI, USA: U.S.; 2004.Google Scholar
- Folkard S, Tucker P: Shift work, safety and productivity. Occupational Medicine 2003, 53: 95–101. 10.1093/occmed/kqg047View ArticlePubMedGoogle Scholar
- Gellerstedt S: Job-rotation rosters for mechanised operations (LIRO Report Volume 22, Number 10).. Logging Industry Research Organisation, Rotorua, New Zealand; 1997.Google Scholar
- Golsse J-M: Effects of shift length on hourly productivity: a preliminary review. Canadian Forest Industries 1991, 34–39. (1991: Aug/Sept).Google Scholar
- Hanna AS, Taylor CS, Sullivan KT: Impact of extended overtime on construction labour productivity. Journal of Construction Engineering and Management 2005,131(6):734–739. 10.1061/(ASCE)0733-9364(2005)131:6(734)View ArticleGoogle Scholar
- Kerin A, Carbone J: Financial opportunities in extended hours operations: Managing costs, risks, and liabilities. Circadian Technologies, Lexington, MA, USA; 2003.Google Scholar
- Kirk P: The impact of shift length on processor operator fatigue (LIRO Report Volume 23, Number 18).. Logging Industry Research Organisation, Rotorua, New Zealand; 1998.Google Scholar
- Maxwell D: A production analysis of the night shift in a forest operation. B. Sc. Thesis. University of New Brunswick, Canada; 1982.Google Scholar
- McKerchar A, Twaddle A: Central log yards – an option to consider (LIRA Technical Release. Volume 9, Number 2).. Logging Industry Research Association, Rotorua, New Zealand; 1987.Google Scholar
- Mitchell D: Extended working hours in the Southeastern logging industry. PhD thesis. Auburn University, Auburn, AL, USA; 2008.Google Scholar
- Murphy G: Worldwide experience with mechanization and value recovery. In Wood for Africa 2002 Conference, Pietermaritzburg, South Africa. Oregon State University, College of Forestry, Corvallis, Oregon; 2002.Google Scholar
- Murphy, G, & Vanderberg, M. (2007). Modelling the economics of extended shift and 24/7 forest harvesting. New Zealand Journal of Forestry, 14–19.Google Scholar
- Nevison J: White collar project management questionnaire report. Oak Associates, Concord, MA, USA; 1992.Google Scholar
- Nicholls A, Bren L, Humphreys N: Harvester productivity and operator fatigue: working extended hours. International Journal of Forest Engineering 2004,15(2):57–65.Google Scholar
- Passicot, P, & Murphy, G. (2013). Effect of work schedule design on productivity of mechanised harvesting operations in Chile. New Zealand Journal of Forestry Science, 43(2). http://www.nzjforestryscience.com/content/43/1/2. Accessed 4 November 2013.Google Scholar
- Rose K, Visser R, Guild B: Analysis of shift pattern effect on productivity at the Kaingaroa Central Processing Plant. Austro2007/FORMEC'07: Meeting the Needs of Tomorrows' Forests: New Developments in Forest Engineering, Vienna, Austria; 2007.Google Scholar
- Vernon HM: Industrial fatigue and efficiency. George Routledge & Sons, Ltd., New York, USA; 1921.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.