STRUCTURAL ANALYSIS OF ELECTRICITY CONSUMPTION BY MANUFACTURING SECTORS. THE SPANISH CASE Vicent ALCÁNTARA Universitat Autónoma de Barcelona Pablo DEL RÍO Consejo Superior de Investigaciones Científicas (CSIC). Félix HERNÁNDEZ Consejo Superior de Investigaciones Científicas (CSIC) Venice, June 19th 2009

INDEX Aim Methodology Main results The sector-based approach Policy implications Limitations.

OVERALL AIM The aim of this paper is to identify the “key” sectors that contribute most to electricity consumption in Spain. To infer policy recommendations.

METHODOLOGY Based on input-output tables. Complemented with a sector-focused study. For each “key” sector, we identify the availability of electricity-efficient technologies, the barriers to their adoption and the policies to increase EE in those key sectors.

METHODOLOGY Our input-output approach allows us to link a vector of direct electricity consumptions to a symmetrical input- output table good by good… …to show the structural relationship between productive activities and final electricity consumption… …in order to identify the key sectors in electricity consumption in Spain.

METHODOLOGY The good by good input-output table has been obtained from the 2004 origin-and- destiny tables published by the National Statistical Office (INE) in Spain. The information on electricity consumption stems from OCDE Energy Balances (2008).

METHODOLOGY The impact of supply and demand on electricity consumption is identified Four relevant effects: Pure backward. Own backward. Pure forward. Own backward.

METHODOLOGY OWNPURE BACKWARDHow variations in the final demand for the products of a sector affect electricity demand in the sector itself i (demand) → i (EC) How variations in the final demand for the products of a sector affects electricity consumption in other sectors. i (EC) ← j (demand) FORWARDHow variations in the production of a sector affect electricity consumption in the sector itself i (production)→ i (EC). How variations in the production of a sector affect electricity consumption in other sectors i (production) → j (EC) EC = Electricity consumption

MAIN RESULTS Key sectors: 62% of direct electricity consumption 1Agricultural products 12Coke and oil refining 13Production and distribution of electricity 32Basic chemical products 34Pharmaceutical products 35Other chemical products 42Metallurgy 43Metal products 46Other machinery 65Trade and repair of motor vehicles. 67 Wholesale trade 68Retail trade 71Rail transport services 82 Telecommunication services 86Real-state services 95Technical consultancy in architecture and engineering. 100Public administration 102Education services.

MAIN RESULTS Plotting key sectors on the demand side Retail trade Public administration Real-state services Wholesale trade Metallurgy Electricity generation

MAIN RESULTS Plotting key sectors on the supply side Metallurgy Electricity generation Wholesale trade Metal products Real-state services oil refining Retail trade Public administration

MAIN RESULTS In order to reduce electricity consumption, policy measures should be adopted in sectors with a high: own backward effect own forward effects pure forward effects. In contrast, measures to reduce electricity consumption would not be effective in sectors with a high pure backward effect because these sectors are not directly responsible for electricity demand.

MAIN RESULTS However, the results of the exercise do not allow us to identify which instruments should be applied in which sector, i.e., it does not allow us to select a specific measure.

MAIN RESULTS Thus, we have complemented this analysis with an identification of the technologies which could reduce electricity consumption in the key sectors. For those sectors with high pure backward and forward (own and pure) effects: we identify the available technologies, the barriers to their uptake and the policy measures (if needed) to encourage their development and/or adoption.

THE SECTOR-BASED APPROACH The methodology is based on an extensive overview of the (international) literature on technologies which reduce energy use or CO2 emissions in the sectors. We have tried to identify the technologies would be more suitable to reduce electricity consumption in Spain.

THE SECTOR-BASED APPROACH We rely on secondary sources and, particularly, the background sectoral analysis carried out in the context of the Spanish Energy Efficiency Strategy (E4). Some caveats. The main barriers to the adoption of these technologies are further identified in each sector. The instruments to tackle those barriers are discussed.

THE SECTOR-BASED APPROACH Example: Electricity-efficient technologies in the iron and steel casting sector. CategorySpecific technologies Horizontal 1 Improvements in lighting 1.1 Metering and control of electricity consumption per area Improvement in lighting equipment (replacement by energy- efficient lamps, i.e., fluorescent lamps). 2 Regulation and control 2.1. Operation and maintenance: 2.2. Improvement in heating conservation, avoiding leakages when opening furnaces. Appropriate pressure in furnaces and appropriate insulation. Production processes Improvements in furnaces.

POLICY IMPLICATIONS i) The intensity of the backward and forward (own and pure) effects suggests that some measures might be more appropriate than others: In those sectors with a high own effect (backward and forward), sector-specific measures are more appropriate In those with a high pure effect, cross- sectoral measures are more appropriate.

POLICY IMPLICATIONS Identifying the most appropriate measures according to relevant effects… OWNPURE BACKWARD Internal measures, i.e., to increase the own electricity efficiency of the sector (sector- specific measures). Electricity generation, Metallurgy, retail trade, public administration, Measures should aim at reducing the electricity demand of the “j” sector. Metal products, other machinery, wholesale trade, retail trade, real- state services FORWARD Internal measures, i.e., to increase the own electricity efficiency of the sector (sector- specific measures). Oil refining, Electricity generation, Metallurgy, public administration, Measures to reduce electricity consumption associated with the supply of sector “i”. Electricity generation, metal products, wholesale trade, real- state services real-state services

POLICY IMPLICATIONS ii) Instrument combination. We need to combine instruments which reduce end-use electricity consumption (scale factor) with other measures aimed at improving the specific electricity-efficiency (kWh/ton of product) of end-user sectors (technological factor).

POLICY IMPLICATIONS iii) Need to apply sector-specific and cross-sectoral measures. Instruments need to be implemented in order to increase EE in the power generation sector and in the key sectors. These are highly diverse sectors and, thus, the adoption of measures should take their technoeconomic particularities into account. A one-size-fits-all solution is at odds with the inherent technoeconomic complexity of each sector.

POLICY IMPLICATIONS iv) Cost pass-through onto electricity prices should be favoured. The price signal is crucial to encourage EE investments. Climate policies and higher fuel demand would lead to higher energy prices. Ensuring that such price increases are passed on to consumers is important to promote the uptake of EE technologies. Regulations preventing such cost pass-through and maintaining artificially low electricity prices are one of those barriers. Thus, governments should remove regulations which do not allow variations in wholesale electricity prices to be reflected in retail prices in order to encourage EE investments.

LIMITATIONS The costs of reducing electricity consumption are disregarded and, thus, we are unable to say anything about the economic feasibility of reducing electricity demand in a given sector. Cost- effectiveness is a crucial variable for designing policies. Information on the technologies, barriers and instruments that should be considered can only be provided by an additional bottom-up analysis. EE policies should be based on this in-depth analysis of sector-level data.

AN IMPORTANT LESSON… Given the pros and cons of different economic approaches and methods to the analysis of electricity efficiency, their combination would be preferable to the use of individual approaches. These approaches may include input-output tables, transaction cost economics, neoclassical and evolutionary economics…

SOME QUESTIONS (for the audience) What type of policy recommendations can be derived from the input-output approach? Can we go beyond the general policy implications? Can we say something about the type of instrument to be implemented? Does the sector-based approach make sense? i.e. what is its value added with respect to the input-output approach? Are the policy-implications derived from the input-output analysis enough (i.e., need to tackle the key sectors)?

¡¡THANKS FOR YOUR ATTENTION!! Pablo del Río, CSIC ?