The amount of energy needed to manufacture a photovoltaic cell depends on the cell technology in question. A study by Murphy & Spitz from November 2009 found that polycrystalline and monocrystalline cells in particular consume more energy than thin-film cells do, as the processes used to produce the silicon are more complex. However, this rule doesn’t hold if the crystalline cells are produced using the string-ribbon method, which involves drawing two wires through molten silicon to form a long ribbon of silicon. Because this method does not require wafers to be cut from an ingot, there is no material loss of the kind that you get with the other cells. Producing string-ribbon cells consumes about the same amount of energy as thin-film cells do.
When assessing how much energy it takes to manufacture a solar cell, it is important to also consider the energy payback time. This shows how long (in years) it will take the system to produce enough energy to offset the amount that went into its production. As well as taking all system components into account, the figure also includes the amount of energy that will go into recycling them. According to the 2011 EPIA Sustainability Working Group Fact Sheet, solar cells that are located in Germany and receive average annual global horizontal irradiance will have an average energy payback time of roughly two years. This means that a solar cell with a minimum lifetime of 20 years will produce ten times more energy than was needed to manufacture it in the first place.
This question was answered by Robert John Doelling of www.energie-experten.org, an online magazine specialising in all matters related to energy.