Monitoring and Control of Fuel Usage


The cost of fuel amounts to 30-35% of total vehicle operating costs for freight and passenger transport in South Africa. Fuel prices have increased dramatically over the past 5 years and will increase further in future. Effective monitoring and control of fuel usage is an essential part of transport operations management requiring a focused, systematic and comprehensive approach which treats fuel in the same way as control of money. Fuel is a highly saleable commodity and it is very easy to mask pilfering in the absence of effective control systems.

A considerable proportion of transport companies in South Africa do not have systems that can provide them with the potential for fuel saving through improved monitoring and management of the fuel usage of their fleets. In most companies, systems exist to account for all fuel purchased and in most of them the distribution of the fuel to individual vehicles is recorded and average fuel consumption is calculated. This process is analogous to a supermarket clearing all tills weekly and averaging and totalling the takings before recording them into the books. Many transport operators do not have systems that treat fuel as money.

For effective control, all fuel purchases should be authorised, recorded and entered into a control system which can be manual in small operations or electronic for larger volumes and fleets. The system must permit capture of every transaction daily if control is to be maintained. The records of fuel usage must then be related to vehicle usage in a daily monitoring system in order to permit queries and follow up action with drivers. It is futile to engage with drivers regarding average fuel consumptions incurred in a previous week or month; no one can explain a variance older than 48 hours with any certainty even if driving history or logbook is available. Ineffective follow-up promotes abuse of the system.

The basic problem with monitoring fuel consumption of vehicles is the very rapid changes and therefore the variability of the consumption minute by minute or mile by mile. Fuel consumption is not precisely linear with distance, due to the influences of, speed, load, acceleration, terrain, vehicle condition and several other driver related factors.

In order to monitor fuel consumption of vehicles, and particularly freight vehicles, for the purpose of taking managerial action to improve it, the consumption must be monitored for every fill-up of fuel (and where possible all fill-ups should fill the fuel tank). Raw fuel consumption data is ineffective for control purposes due to the variability created by the abovementioned factors. To overcome this, the consumption recorded must be analysed statistically using a methodology that is capable of isolating  variances from the desired standard consumption.

There are numbers of “fuel management’ systems based on electronic technology ,badge readers, RFID, sensors etc which can be effectively used as tools, but in the final analysis it is still essential to find a way for fuel consumption by each driver to be reported and analysed. This must result in a daily report that permits appropriate action to be taken in the event of variations from defined standards for each vehicle or vehicle group. The reporting system must reduce the complex data to a simple but accurate single indicator that action is required; busy managers do not have time to scroll though sheets of data, they need answers not questions.

The average transport operations manager does not have the time or inclination to perform detailed analyses of each vehicle’s performance on each tank of fuel, but that is in fact the only way to establish whether the usage is within standard performance parameters or not.  In most fleets there is the usual “bell curve” of fuel consumptions with 20% below average and 20% above average. Repeated trials show that if the vehicles and drivers in the fleet, that give substandard performance can be trained, controlled or fixed, the savings can be as high as 5-10 % of the total fuel bill.

If operating standards are tightened including, driving at optimal engine revolutions, reduced top speeds, reduced idling, improved vehicle maintenance and better vehicle specifications and some modifications, there is often further potential saving of 5-8% in even well run fleets, with minimal change in fleet performance. In most fleets surveyed on roller dynamometers, even fairly new vehicles are typically 15-20% down on power relative to manufacturer’s specifications.

Due to the normal fluctuation of fuel consumption in heavy vehicles the major problem is to identify the significant variations and to have systems that provide effective identification of where to take immediate action. The analytical system must also be able to identify immediately, whether the remedial action was effective. It is possible to do such analyses and there is a need for systems training in this area. It can be predicted that progressive fleets will see the need for a new breed of more dedicated “fuel control” clerks to get an effective handle on this massive slice of company expenditure.

There is a statistical technique for doing this analysis and it is relatively simple, but is very seldom employed due to the lack of an appropriate system to produce manageable data that can be used for managerial action. The technology is available, to analyse the information that is currently gathered manually or electronically, and often presented in sophisticated programs. As a consultant, from many company visits, it is rare to find a manager that has any serious, continual action programme arising out of his fuel monitoring system, thereby partly negating its usefulness.

The solution presented by use of the Cusum (cumulative sum of the variance) statistical technique can be simply explained by reference to a typical fuel record sheet as shown in table and graph illustrated below.

The table shows the typical daily variations in fuel consumption for a small truck working in an urban area, sometimes loaded, part loaded, over different gradients at varying speeds on streets and freeways. The absolute values are irrelevant as the system works at any fuel consumption level.

The daily fuel consumption in litres per 100 kilometres fluctuates as shown by the blue line in the following graph. The variations make it almost impossible to decide whether or when the fuel consumption was in line with the standard for the class of vehicle (indicated as K=10 litres per 100).

The daily fuel consumption (L/100 Kms) is very simply converted to Cusum by the formula; L/100 Kms – K

The next step is to add the variance for each fillup sequentially, to give a continuous cumulative sum of the variances (“cusum”, in right hand column). As shown in the graph, the cusum takes on a very different shape, and the slope of the line, indicates immediately that consumption is out of control. If consumption was under control the cusum line would be horizontal. Note: the actual position of the line is irrelevant; it is the angle of the slope that indicates the variance.

The power of cusum is that it:

  • gives an instant visual check,
  • preparation can be done by any clerk, manually or electronically.
  • it shows over-consumption for each fill-up, permitting immediate investigation
  • it shows the magnitude of the variance
  • it defines steady over-consumption compared to spikes indicating theft
  • If a correction is made e.g. calibrate fuel pump, change driver; the effect is immediately visible.

The cusum technique offers one of the most effective tools for management of fuel consumption, and can be relatively simply programmed into a computer to provide instant daily control.

This is the second of a series on Road Transport Management by  For queries, assistance or discussion contact