The term 'horse power' is largely credited to James Watt, in the late 1700s. Watt was a Scottish engineer who invented a number of improvements to steam engines, which he then (in partnership) began to manufacture and sell (the first units going into service in 1776).

Most of Watt's potential clients were using horses, so he soon found that in order to market his engines, he needed to express the power of his engines in terms of how many horses a given engine would replace. For this purpose, he first calculated the average power of a horse, which he termed 'horsepower'. He then specified for each of his engines how much 'horsepower' it had (i.e. of how many horses it had the equivalent power of). Due to the success of his steam engine business, the term 'horsepower' came to be recognized. His competitors and other machine manufacturers copied his approach, also expressing the power of their engines in 'horsepower', which resulted in it becoming a standard measure of power.

To calculate the force of a horse, Watt first observed that a common use of horses was to walk in a circle, pushing a beam attached to a mill wheel (see following illustration). He then noted that the circle that the horse would walk was typically about 12 feet in radius, so the distance around the circle (circumference) would be about 75 feet (2.4 × 2π × 12). He also measured that a horse would go around the circle an average of 144 times per hour and he judged that as the horse walked it pulled with a force of about 180 pounds. Multiplying this out and then rounding off, he ended up with a figure of 33,000 foot-pounds per minute.

Horse Power Calculation

Although Watt was the first to widely publicize a measure of horse power, he was not the first to come up with the idea. In 1702 (almost a century earlier), Thomas Savery wrote in The Miner's Friend about the idea of expressing the power of engines in terms of horse equivalents. However, it was Watt who first took the idea and introduced it into widespread use.

Other people have measured the power of horses and ended up with somewhat different figures; John Smeaton estimated it at 22,916 foot-pounds/minute and John Desaguliers produced a figure of 27,500 foot-pounds/minute. However, the figure produced by Watt was accepted as the standard definition of horsepower.

Alternative Histories

Some historians have disputed the details of how Watt calculated the power of a horse. For example, some say that he actually calculated the power of a pit pony (a type of pony commonly used in UK mining) and then added 50%  to determine the power of a horse. Consequently, the exact method which Watt used to calculate the value of a 'horse power' is open to dispute, but the remaining historical details are well documented and accepted.

Horse Equivalents

The actual power which a horse can exert depends largely on the duration over which the horse must work. Watt's calculation was based on a 4-hour working shift, where the horse would then rest before working again. If a horse works for very short periods, it can exert higher levels of force. For example, a horse can exert the equivalent of up to 15 horsepower, if the effort is just a few seconds. Alternatively, if a horse has long working shifts, the amount of force it exerts over this longer period will be less.

On the other hand, a machine works continuously, producing the same level of power at all times. So, a 1-horsepower machine working 24 hours a day will do 3 times as much work as a horse working two 4-hour shifts. Consequently, taking into account the working hours of a horse per day, a machine which is in continuous use would be equal to about three times as many horses as its specified horsepower would indicate. For example, in this case a machine of 2 horsepower would in fact replace about 6 horses (each of which work an average of 8 hours per day).

Note also that the term horsepower is an approximate calculation for a working horse. Depending on various factors (race, health, size, condition, age, application), the actual power and working capability of the individual horse could be higher or lower than 1 horsepower.

Related articles