skip to primary navigation skip to content

The History of Fuel in England, c.1550-1850

The History of Fuel in England, c.1550-1850

A pilot project

In autumn 2019 we undertook a pilot project to test and refine a method of model-building aimed at identifying the market areas of commercially available fuels in early modern England, during the period of the transition to coal as the dominant fuel (Warde, 2006). The pilot project worked with the simple hypothesis that price was a significant determinant of fuel choice, and that without major shifts in the technology of mining before 1800, changing transport costs would be a major (perhaps the major) determinant of the spread of coal use. If one could establish freight costs at a given date, then in principle one could establish a national map of implied coal prices. This could then be compared with data of the actual prices consumers were prepared to pay, in theory giving a rough indication of where the 'frontier' of affordable coal was at a given point in time. Of course, coal prices were by no means solely determined by freight costs in reality, for once they got beyond the pithead they could be subject to other charges such as tolls and taxation.

Fortunately, a state of the art high resolution GIS model of freight costs across England for 1680 and 1830 exists and we were generously given permission to use it (Alvarez et al, 2017). We used the freight price model to develop a pilot coal price model. The purpose of our pilot model was not to develop the perfect model, and nor do we think such a thing would be possible. Instead, we wanted to (1) identify the regions where the 'coal frontier' might be at particular moments in time, and (2) test the cost assumptions of the model against empirical price data to see if there were consistent deviations. This would allow us to identify the most interesting localities for future research. In addition, we could also work with the model to see if it could help us understand the changing costs of coal as the transport network and other factors changed in the years between 1680 and 1830.

The model employed worked by linking standard flat-rate transport and transhipment costs (by road, navigable waterway, and coastal shipping) to a multi-modal transport network in GIS incorporating the road network ,waterways and coastal shipping routes. The nodes in the model consisted of 590 towns, and the model was used to generate a matrix of all the costs of all possible transport routes between each town node in the system. The cheapest route would, most likely, represent the route any freight was most likely to take, and those lowest-cost routes could then be used to establish a likely minimum price for coal at each node.

It was established which nodes in the model fell within those areas exploited as coalfields at the benchmark date of 1680, and this was then used to produce national maps of estimated coal prices. It should be remembered that these are prices at nodes in the transport network which may be significantly lower than in nearby localities because of the high costs of overland freight. To come to an estimated price of coal per ton, transport costs for each town were added to estimated baseline coal prices within the coalfields. For ease of interpretation, however, the coal prices per ton were then estimated for the whole of England and Wales using inverse distance weighted spatial interpolation. In reality, we would expect prices to be higher at more difficult to access rural locations, and hence a true price map to be more complex. The standard assumption of contemporary commentators and historians is that coal carried by land had doubled in price every ten miles, but this must have varied considerably with location.

This method is indicative rather than exact, and we are well aware that the coal and freight price data available may contain considerable uncertainties because of variations in coal quality, demand and weights and measures. The model itself is, however, a useful means of testing the interpretation of this data.

The initial results for c.1680 are shown in Map 1.Map 1. estimated coal prices for England and Wales c.1680

Our thanks go to Eduard Alvarez, Dan Bogart and Leigh Shaw-Taylor for permission to use their freight cost model without which we could not have produced this map.

This map shows that we would expect the range of coal prices found nationally to fall between the baseline estimate of 70 pence per ton within the coalfield and more than eight times that much for inland areas lacking navigable waterways and far from any coalfield. The error in rural prices is likely to be higher in the north, because of the lower density of towns, and thus the greater distances, often over difficult topography, to more remote places. A revised version of the model would need to add more places with estimated transport costs in these types of areas. In southern England where towns are much denser the map shows a stark difference in estimated price short distances beyond the coast and navigable rivers. Within the Ouse catchment estimated prices are in the vicinity of 200d per ton near the Ouse and its tributaries but climb steeply landward and double within 10 miles in some places such as east Norfolk.

These results can be compared with the first moment when a national map of coal prices based on empirical data can be produced, from a survey of 480 Poor Law Unions in 1842-3 that included the local costs of coal (Map 2).Map 2 Estimated coal prices 1842-3

By 1843 we would expect to see considerable changes due to the extension of the navigable waterways since 1680, although the development of the railway network had yet to make a major impact. Similarly to the method applied to the earlier model, coal prices were estimated for the whole of England and Wales using inverse distance weighted spatial interpolation, with the same caveats to be applied. It is notable that while the overall peaks of estimated price beyond 500 pence per ton are no longer reached the coal price frontier in parts of the south does not appear to have shifted much in a century and a half. Prices stayed stubbornly high in much of Hertfordshire, parts of Essex, Surrey, Sussex and Kent.

These findings suggest that although as we would expect the range of national prices had narrowed, the national gradient of prices had remained high. Coal had become relatively cheaper, but coalfields retained a considerable cost advantage, and the band of high fuel prices running from inland East Anglia west and south-west across the southern Midlands and into the south-west peninsula remained unchanged. This gives us a fairly stable region for the investigation of transition across the period which would of course be partly determined by the availability and cost of other, 'traditional' fuels. Nevertheless, it is also clear that our estimated coal prices are relatively stable across the two dates, despite more general inflation. Hence the real cost of coal must have fallen everywhere. The situation with more traditional fuels in each locality is uncertain. Rents and other input costs must have risen, but real labour costs in the countryside had fallen from the mid-eighteenth century.

Of course, a major question is whether the estimated prices are correct. An ongoing area of work in the pilot project has been to compare these with empirical prices obtained from both secondary sources (e.g. Nef, 1932, Beveridge, 1939, Hatcher, 1993, Langton, 1979) and archival research especially in Cambridgeshire, Suffolk, and Bedfordshire. For some places (for example, Cambridge and London) we have price data for nearly every year across the period. Nevertheless, for others, even individual spot prices allow us to identify 'deviations' from the model. For an indication of the geographical breath and chronological depth of this coal price dataset see the video of prices in Map 3.

Direct comparison with the estimated prices for 1680 derived from the modelling exercises above is rendered difficult by the scarcity of prices before 1842-3. Despite over 2500 individual prices being entered in the database and mapped in the GIS (and this sample continually expanding), at present the best year (1692) only has 42 distinct locations with documented prices. This and the large range in prices in the same location exhibited only 20 years either side of 1680 also renders the exercise problematic. It seems that prices could be subject to considerable short-term variation, itself an important consideration in understanding transitions. This is not due to errors in the price data. Seaborne coal prices fluctuated due to the Anglo-Dutch war 1672-4, the Nine Years War 1689-97, Queen Anne's war 1702-1713 as well as taxes principally the 5s per ton King's Duty tax from1698. Current testing of the model suggests that raster values are generally too high when compared with real prices. However, the model serves as a useful as guide to relative price differences from one region to another.

The pilot model also allowed us to reverse the question. Instead of asking what the price of coal might be at any node in the system, we could constrain the model according the maximum price we found any consumer was prepared to pay in particular areas of the country. By using localised minimum (pithead) and maximum prices, we could estimate a maximum supply zone for any given coalfield, and also the price gradient within that supply zone.

This helps us to identify the likely source of coal at any given place, something that is not immediately apparent using the national model. Importantly, this supply zone can be linked with the range of estimates of coal supply from particular coalfields in the late seventeenth century (Hatcher, 1993). This in turn can be linked with the country population estimates for 1700 produced by Tony Wrigley (2009). Although we are obviously dealing an uneven distribution of both population and coal supply within these zones, and their boundaries do not coincide, this allowed rough-and-ready estimates of coal consumption per capita.

Map 4 presents the results of this exercise for the Staffordshire coalfield.Map of estimated coal prices from the Staffordshire coalfield c.1680

It shows that before the opening of the Staffordshire and Worcestershire Canal in 1777, the supplies of the Staffordshire must have been consumed within a relatively narrow region largely stretching southwards into the Birmingham region and the middle Severn valley. It revealed, for example, that the figure for the Staffordshire coalfield was probably twice as high as that for London, and very much than those for other regions in southern England. It raises the question of whether this difference can be accounted for by differences in fuel use among households, industries, or both.

The pilot work demonstrates the possibility of building up a much more fine-grained understanding of fuel availability and its significance in early modern England, in a period before the dramatic expansion of coal use in the smelting of iron (predominantly after 1780) and to generate motive power in steam engines. The persistence of a high price gradient in certain regions from the seventeenth to the nineteenth centuries also indicates that one of the major factors in the greatly increased supply of coal was differential rates of consumption between different regions, and not just an expansion of the area where coal was the habitual fuel of homes and industry. On the basis of these results the north Midlands and western Yorkshire have been identified as areas of particular interest for examining the phenomenon of locally intense coal use. The Ouse catchment in eastern England has been identified as a region of particular interest for understanding patterns of consumer choice among different fuels and among different groups of users. This is due to multiple fuel availability (coal, firewood, peat and turf), connection to different zones of coal supply from the Wash, east Midlands and Thames, with persistently steep price gradients for coal at a frontier that can be traced in detail.

Equally, the pilot use of the model has demonstrated the possibility of using it both (1) test the effects of varying key variables (pithead prices, regional transport costs, and the extent of the navigable waterway and road network) and (2) populate it with empirical price and usage data to develop an economic geography of fuel in early modern England. In turn this can be related to new, fine-grained datasets on local populations, occupational structure, and industrial location to examine the relationship between fuel availability, regional development, and resource use more widely.


  • Alvarez, E., Bogart, D., Shaw-Taylor, L., Dunn, O., Satchell, M., 'Growth before steam: A GIS approach to estimating multi-modal transport costs and productivity growth in England, 1680-1830.', Working Paper, 2017.
  • Beveridge, Sir W., Prices and wages in England from the twelfth to the nineteenth century (London: Longmans & Green, 1939)
  • Hatcher, J., A history of the British coal industry before 1700. Vol. I. Towards the age of coal (Oxford: Clarendon, 1993)
  • Langton, J., Geographical change and industrial revolution. Coalmining in south west Lancashire, 1590-1799 (Cambridge: Cambridge University Press, 1979)
  • Nef, J.U., The rise of the British coal industry. 2 vols (London: Routledge, 1932)
  • Warde, P., Energy consumption in England and Wales, c.1560-2000 (Naples: CNRI, 2006)
  • Wrigley, E.A., 'Rickman Revisited: The Population Growth Rates of English Counties in the Early Modern Period', Economic History Review, 62, (2009), pp. 711-735.

Sources for Coal Prices


Account of Prices of Articles of Consumption at Poor Law Unions in England and Wales, June 1842 and 1843, BPP (1843), XLV; Davies, D., The case of labourers in husbandry, stated and considered in three parts (New Jersey, 1977); Eden, F.M., The state of the poor. An history of the labouring classes in England from the conquest to the present period London, 1797); Hatcher, J., A history of the British coal industry before 1700. Vol. I. Towards the age of coal (Oxford, 1993); Tunzelmann, G.N. von, Steam power and British industrialization (Oxford, 1978); Nef, J.U., The rise of the British coal industry. 2 vols (London, 1932); Beveridge, Sir W., Prices and wages in England from the twelfth to the nineteenth century (London, 1939); Langton, J., Geographical change and industrial revolution. Coalmining in south west Lancashire, 1590-1799 (Cambridge, 1979), Thorold-Rogers, J.E., Six centuries of work and labour (London, 1886); Houghton, John. A collection for the improvement of husbandry and trade 3 vols (London, 1727-8); Spufford, M., 'Chimneys, wood and coal', in P.S. Barnwell and M. Airs, (eds), Houses and the hearth tax. The later Stuart house and society (York, 2006), pp.23-32.


Bedfordshire Record Office OE2071/79, P46/12/6,

British Library Add MS 34110; Add MS 33509.

Cambridgeshire Record Office P52/5/1, P58/11/1, P93/12/2, P145/12/1, P62/12/5.

Pembroke College, Cambridge Treasury Accounts.

Norfolk Record Office WLS XIV/17-20, Y Coal Journals, Y Chamberlains' Accounts, ANW/5/6/18, BL/Y/1/11, BL/CS/6/7/15, BUL 11/88/615X8, Mundes Charity, GUN 130/625X5, Sam Browne's Cashbook.

The National Archives C1.575.13, C1.961.46, C1.1476.39, C1.1052.23, C1.1139.44

West Suffolk Record Office FB 79/G1/1-2, FL 562/7/8, FL667/7/2, FB 77/G5/1, FL 501/7/1, FL 528/7/1, FL 555/7/1, FL574/7/10, FL 599/7/4, FB 64/G2/1, FB 80/G1/1, FB 78/G2/3, FL 641/7/1, FL640/7/3, FL514/7/4, FB 65/G1/1, FL 652/7/2, FL 655/7//4.

We would like to thanks Oliver Buxton and Fiona Williamson for their assistance in obtaining coal prices.