Eficiência Energética

ELSEVIER PI1S0301421596000171 Energy Policy Vol 24 No 5 pp 377390 1996 Copyright 1996 Elsevier Science Ltd Printed in Great Britain All rights reserved 0301421596 1500 000 What is energy efficiency Concepts indicators and methodological issues Murray G Patterson Department of Resource and Environmental Planning Massey University Private Bag 11222 Palmerston North New Zealand This paper critically reviews the range of energy efficiency indicators that can be used particularly at the policy level Traditional thermodynamic indicators of energy efficiency were found to be of limited use as they give insufficient attention to required end use services The specific limitations and appropriate uses of physicalthermodynamic economicthermodynamic and pure economic indicators of energy efficiency are also considered The paper concludes with a discussion of the per sistent methodological problems and issues which are encountered when attempting to operational ize all of the energy efficiency indicators These include the role of value judgements in the construction of energy efficiency indicators the energy quality problem the boundary problem the joint production problem and the question of isolating the underlying technical energy efficiency trend from the aggregate indicator Copyright 1996 Elsevier Science Ltd Keywords Energy efficiency Definitions Indicators Energy efficiency now has an important place in the public policy agenda of most developed countries The importance of energy efficiency as a policy objective is linked to com mercial industrial competitiveness and energy security benefits as well as increasingly to environmental benefits such as reducing CO 2 emissions Despite the continuing policy interest and the very many reports and books written on the topic of energy efficiency little attention has been given to precisely defining the term 1 The purpose of this paper is therefore to open up this debate by critically re viewing the range of possible energy efficiency definitions and how they can be operationalized by the use of indica tors The methodological problems and issues which are then encountered when attempting to operationalize such definitions will also be discussed Energy efficiency is a generic term and there is no one unequivocal quantitative measure of energy efficiency In stead one must rely on a series of indicators to quantify changes in energy efficiency In general energy efficiency IFor example in New Zealand three recent reports on energy efficiency have been released by the Ministry of Commerce Harris et al 1992 the Electricity Corporation of New Zealand ECNZ 1993 and the New Zealand Planning Council Terry 1991 None of these reports however has explicitly defined the term energy efficiency refers to using less energy to produce the same amount of services or useful output For example in the industrial sec tor energy efficiency can be measured by the amount of en ergy required to produce a tonne of product Hence energy efficiency is often broadly defined by the simple ratio 2 Useful output of a process Energy input into a process The issue then becomes how to precisely define the useful output and the energy input which in turn gives rise to a number of important methodological considerations which are often ignored in the literature A number of indicators can be used to monitor changes in energy efficiency These fall into four main groups 2The useful output of the process need not necessarily be an energy output It could be a tonne of product or some other physically defined output or it could be the output enumerated in terms of market prices Energy efficiency indicators sometimes involve ratios that reverse the numerator and the de nominator For example the energyGDP ratio commonly used as an indi cator of energy efficiency eg Wilson et al 1994 is constructed in this way It could be argued by reversing the numerator and the denominator that an energy intensity ratio is now being constructed However as such ratios implicitly contain information about the energy efficiency of a process they are therefore covered in this review of energy efficiency indicators 377 378 What is energy efficiency M G Patterson 1 Thermodynamic these are energy efficiency indicators that rely entirely on measurements derived from the sci ence of thermodynamics Some of these indicators are simple ratios and some are more sophisticated measures that relate actual energy usage to an ideal process 2 Physicalthermodynamic these are hybrid indicators where the energy input is still measured in thermody namic units but the output is measured in physical units These physical units attempt to measure the ser vice delivery of the process eg in terms of tonnes of product or passenger miles 3 Economicthermodynamic these are also hybrid indi cators where the service delivery output of the process is measured in terms of market prices The energy input as with the thermodynamic and physicalthermo dynamic indicators is measured in terms of conven tional thermodynamic units 4 Economic these indicators measure changes in energy efficiency purely in terms of market values That is both the energy input and service delivery output are enumerated in monetary terms Thermodynamic indicators In one sense thermodynamic indicators of energy effi ciency seem to be the most natural or obvious way to meas ure energy efficiency as thermodynamics nowadays is often defined as the science of energy and energy pro cesses Surprisingly though thermodynamic measures of energy efficiency are not as satisfactory measures of energy efficiency as they might at first appear There are good methodological and operational reasons for not whole heartedly accepting the use of thermodynamic measures of energy efficiency which are discussed later on in this paper However one attraction of using thermodynamic quan tities for measuring energy efficiency is that they are calcu lated in terms of state functions of the process This means that they provide unique and objective measures for a given process in the context of a particular environ ment prescribed by temperature pressure concentration chemical formula nuclear species magnetization etc Thus for any change in physical conditions that results from some dynamic process the associated change in the values of the state functions can be uniquely measured or imputed Similarly for a specified change in physical con ditions the minimum energy requirement can be unequiv ocally calculated Firstlaw energy efficiency Firstlaw efficiency is also referred to as thermal efficiency or enthalpic efficiency This is because it measures effi ciency in terms of the heat content of the inputs and outputs of the process and heat content is measured in terms of enthalpic change values A The enthalpic efficiency ratio for any process is there fore the Mvalue of the useful output of the process di vided by the Avalue of the inputs of the process EAH Mout AHin where EaH enthalpic efficiency Mou t sum of the useful energy outputs of a process An Min sum of all of the energy inputs into a process 64 It is important to realize that the enthalpic efficiency indica tor only measures the useful output for example an in candescent light bulb has an enthalpic efficiency of about 6 In this process only 6 of the input of electricity A is converted to light energy with the other 94 being lost to the environment as waste heat If the waste output of any process is added to the useful output of any process the total output then equals the total inputs when the en ergy is measured in enthaipic terms In essence this is an other way of stating the first law of thermodynamics ie that in any conversion process energy cannot be created or destroyed For this reason enthaipic efficiency is often called firstlaw efficiency The use of enthalpic AH measurements of energy does not take account of the quality of energy No distinction is made between high quality energy sources which are more useful and productive and low quality energy sources which are less useful and productive For example one unit of electricity high quality energy is implicitly as sumed to have the same usefulness as solar energy low quality energy Despite this wellknown deficiency of en thalpy A measures with respect to energy quality many analysts such as Sioshansi 1986 and Schurr 1984 still use these measures in macrolevel energy efficiency stud ies Such studies are misleading as they treat different en ergy inputs as being homogeneous in quality terms They are only strictly homogeneous in terms of heat equivalents but not in terms of any sensible systemwide quality meas ure that takes account of other energy end uses apart from heat Secondlaw energy efficiency using work potentials to adjust for energy quality A significant problem with firstlaw energy efficiencies is that they do not take account of the energy quality of the in puts and the useful outputs Therefore if either the inputs or useful outputs of the two processes are of different qualities you cannot meaningfully compare their relative energy effi ciencies You are comparing apples with oranges 3patterson 1983 1993a has proposed the quality equivalent methodology to measure energy quality in complex economic systems where there are many desired end uses of energy apart from just heat What is energy efficiency M G Patterson 379 A number of thermodynamic quality numeraires 4 can be used to convert the input denominator tin5 in the ther mal efficiency ratio to common quality units in an attempt to overcome this problem of energy quality These quality numeraires are based on secondlaw considerations First it has been suggested by the International Federation of Insti tutes for Advanced Study 1974 among others that Gibbs free energy change AG be used to measure the relative en ergy quality of inputs When a process is carried out at con stant temperature and pressure the decrease in Gibbs free energy represents the maximum work that can be done by a process This decrease in Gibbs free energy AG is defined by AG AH TAS where countered in economic production processes they still pre sent a number of fundamental problems First of all work is not the only useful desired energy output in the economy with modem economies having a significant end use de mand for heat Second it is unclear what type of work should be used as the quality numeraire and this is import ant as not all forms of work chemical electrical mechan ical etc are the same or necessarily commensurable with each other If mechanical work is selected as the quality nu meraire as is often suggested there is no sound theoretical basis for this selection as again mechanical work is only one of the many desired energy outputs in modern economies Ultimately therefore it is argued that the call for commensurating energy inputs Zdtin in terms of some work potential still does not provide a rigorous solution to the energy quality problem which is encountered in using first law energy efficiency indicators AG change in Gibbs free energy AH change in enthalpy T temperature AS change in entropy Other work potentials that could be used to commensurate the energy quality of the inputs include exergy and avail able work The difference between available work and Gibbs free energy is that in the former pressure and tem perature refer to the surroundings whereas in Gibbs free energy they refer to the reference state Hence it is argued by proponents of available work that this is a more realistic measurement of work as it takes account of the physical conditions that exist in reality Exergy is a very similar measurement of work potential being defined by Ahem 1980 as The work that is available in a gas fluid or mass as a result of its nonequilibrium condition relative to some reference condition The reference condition most com monly used is sealevel atmospheric conditions which is considered to be the sink for terrestrial energy systems While both available work and exergy might seem to be more appropriate than Gibbs free energy change AG in that they explicitly refer to environmental conditions en Secondlaw energy efficiency ideal limits Another approach is to define energy efficiency relative to the ideal minimum energy required to undertake a task In mathematical terms this ideal efficiency can be defined by the following ratio P EAHactualfAHideal where P EAHactual EAHideal secondlaw efficiency of a process in per forming a specified task actual enthalpic efficiency of a process in performing a specified task ideal enthalpic efficiency to perform a task reversibly by a perfect device This ratio can therefore be used to measure how close a real world energy conversion process is to the ideal efficiency where the most efficient process possible has an efficiency of 9 1 Often the Kelvin formula for the conversion of heat to work is used in these calculations MAHt I t2tl 41t has also been suggested in the thermodynamic literature eg Groscurth et al 1989 Horsley 1993 that temperature is an appropriate quality nu meraire The rationale for using temperature as a quality numeraire seems to relate to the Kelvin formula which sets the upper limit for a Carnot en gines conversion of heat to mechanical work According to this formula temperature differences between the heat source t I and sink t 2 define the maximum efficiency of converting heat to mechanical work In essence therefore using temperature in this way is tantamount to using maximum mechanical work as the quality numeraire Hence the follow ing discussion in so far asit relates to using mechanical work potentials as quality numeraires also applies to using temperature as a quality numeraire 5The literature and hence this discussion focuses on the commensuration of energy inputs AH inputs in terms of their energy quality However the energy quality problem also occurs when you try to compare two pro cesses with different outputs The outputs also need to be commensurated in terms of their energy qtmlity to enable the valid comparison of the relat ive energy efficiencies of the two processes where M mechanical work done by a conversion process J AH heat input into the conversion process J t 1 temperature of the heat input into the conversion process K t 2 temperature of the heat output from the conver sion process K Temperature differences between the heat source tl and the heat sink t2 therefore limit the efficiency by which heat can be converted to mechanical work Similar temper ature defined potentialities can be shown to quantify the ideal level of conversion efficiency between other sources 380 What is energy efficiency M G Patterson and end uses of energy besides the standard conversion of heat to mechanical work Often 1978 For example by using the Kelvin formula the maximum enthalpic effi ciency of converting heat to electricity can be calculated at EAHideal 712 q 1000 K t 2 288 K and this can be compared with an actual enthalpic efficiency of New Zealand power stations of about 30 Hence the second law efficiency p in this instance is 30712 42 Secondlaw efficiency P measures can be applied to a wide range of processes including chemical Gyftopoulos et al 1974 Sussman 1977 transport Berry and Fels 1973 heat transfer Bejan 1980 Kay and Scholenhls 1980 refrigeration air conditioning and electric drive A practical but not theoretical problem with establishing the ideal minimum energy requirements of processes is that for some processes it is not exactly clear if such calculations can be carried out in an unambiguous fashion yielding a unique result Jaynes 1989 Frequently the strategy is to adapt the Kelvin formula to processes that do not have mech anical energy as their output eg to the process of electricity generation as explained above or to processes that cannot be strictly considered to be heat engines In other circum stances various other methods can be used to define the minimum energy requirements of a task For example Slesser 1982 cites the use of Betzs theory to determine the maximum efficiency of a wind turbine While secondlaw efficiencies based on defining the ideal limits of processes are useful in pointing to the the oretical energy savings that can be achieved by engineering and technical improvements they are restricted in their ap plicability to real world systems The first limitation of the method is that it fundamentally assumes perfect reversibil ity which is equivalent to assuming infinitely slow pro cesses Obviously real world processes are required to occur in finite time periods a chemical engineer requires a chemical reaction to take place within a specified time period if it is to be of any economic value and all engines in actual operating conditions must consider human impa tience which in turn introduces a whole series of unavoid able losses such as friction losses Andresen et al 1977 and Wu I 988 have however de veloped optimization methods to overcome this assumption of perfect reversibility infinitely slow processes which is used in the calculation of ideal energy efficiencies This method termed finite time thermodynamics contains a minimum set of constraints that engines can accept It can be argued that by moving away from infinite time classical thermodynamics Andresen et al 1977 have explicitly ac cepted that socalled subjective factors such as human im patience are of importance in calculating energy efficiency The usefulness of their method however is that it makes explicit the trade off between time constraints and energy use The second limitation of the ideal limit method of en ergy efficiency definition is that it is not capable of taking account of indirect energy inputs Van Gooi 1980 makes this point by citing a number of examples including for in stance the case of increasing the length of a heat exchanger to recover a higher fraction of available heat Usually but not always there is an optimum point between increasing capital equipment to save process energy and the energy lost in the indirect energy embodied in the extra capital equipment The ideal limit method is incapable of con sidering such factors Essentially by including indirect en ergy inputs the energy quality problem is once again encountered as almost inevitably there will be a multipli city of different types of energy inputs that need to be somehow equivalenced Physicalthermodynamic indicators 6 One criticism of traditional thermodynamic indicators of energy efficiency is that they do not adequately encapsulate the end use service required by consumers in the output measurement That is the numerator in the thermodynamic efficiency ratios measure either heat content in the first law efficiency or some work potential in secondlaw effi ciencies Consumers of course do not value the end use service on the basis of its heat content or work potential Therefore energy analysts have developed efficiency ratios that measure the output in physical units rather than in ther modynamic terms These physical units are specifically de signed to reflect the end use service that consumers require For example the desired output of freight transport is the carriage of a given mass of freight over a given distance this output can therefore be measured by tonne kilometres Hence an appropriate energy efficiency measure for freight transport could be Output tonne kilometres Energy input AH One advantage of using these physical measures is that they can be objectively measured just as thermodynamic meas ures can but they also have the added advantage that they directly reflect what consumers are actually requiring in terms of an end use service Because they are physical meas ures these can readily be compared in longitudinal time series analyses That is difficulties are not encountered in time series studies as happens in the use of economic indi cators of energy efficiencies due to changes in market val ues A tonne kilometre or a tonne of product is always a tonne kilometre or a tonne of product whereas the market value of a tonne kilometre or tonne of product can change quite markedly over long time periods If hybrid physicalthermodynamic measures of energy efficiency are to be used it is appropriate that they be devel oped on a sectoral basis as different sectors tend to have different industry based standards for specifying their outputs In the residential and commercial sectors the most frequently 6purely physical indicators of energy efficiency can also be developed eg litres of fuel oiltonne of butter These purely physical indicators are quite limited for comparative purposes as they can only validly be used for records which have the same units for the denominator and numerator used measure is energy inputsquare metre although this does present several problems when it is used to measure the aggregate energy efficiency performance of buildings Accordingly the energy inputsquare metre indicator is some times adjusted to take account of degree days as a signi ficant proportion of energy use in buildings involves space heatingcooling and to take account of hot water usage The Joint Economic Committee of the US Congress 1981 sug gested that cubic metres are a better measure than square me tres although such data are difficult to obtain from official statistics The fundamental problem with either the energy inputm 2 or energy inputm 3 indicator is that they are predi cated on the idea that the main services delivered to build ings are HVAC and lighting and that these are directly pro portional to square or cubic metres Building structures particularly residential buildings are the focus for the deliv ery for many other energy services eg water heating cook ing and mandatory electrical services Hence for the resi dential sector it may be appropriate to also develop indicators for measuring the efficiency of delivery of cook ing and water heating services eg 1 energy inputcooking heat delivered to a specified tem perature 2 energy inputwater heating delivered to a specified temperature Different types of physicalthermodynamic indicators can be developed for the transport sector The output measure ments need to reflect the objective of the specific type of transport activity For freight transport an appropriate indi cator is therefore energy inputtonne kilometres as the function of freight transport is to move a freight mass mea sured by tonnes over a given distance measured by kilo metres For passenger transport energy inputpassenger kilometres or energy inputvehicle kilometres may be ap propriate indicators of energy efficiency It has been sug gested by Collins 1992 that energy inputvehicle kilometres is an inappropriate indicator as the objective of passenger transport is to move people across distances not to move vehicles which may be near empty across dis tances It can also be argued that for many transport opera tions the objective is not tonne kilometres or passenger kilometres but rather tonne kilometres or passenger kilo metres per unit time This is because speed and the neces sity to minimize transport time is the essence of much freight and passenger movement Therefore it could be ar gued that transport energy efficiency indicators should be adjusted to take account of this speed objective which is ap plicable to many transport operations Due to the relative heterogeneity of both the industrial and agricultural sectors in terms of the very different prod ucts produced by various industries any attempt to devise an aggregative physical output measurement is futile For most industries the product can be measured in terms of its mass eg tonnes of butter tonnes of bricks tonnes of wheat tonnes of aluminium Hence appropriate indicators may be What is energy eJiciency M G Patterson 381 1 energy inputtonne of butter 2 energy inputtonne of bricks 3 energy inputtonne of wheat 4 energy inputtonne ofaluminium For other industries volumetric output measurements may be appropriate eg litres of milk cubic metres of wood or timber litres of oil In each case the standard industry measure needs to be applied and care must be taken in pre cisely defining the output eg some industries use oven dry tonnes to measure output rather than tonnes that are inclu sive of water content The measurement of energy efficiency in terms of phys icalthermodynamic indicators is not as straightforward as it first appears because of the socalled joint production or partitioning problem This refers to the difficulty in allocat ing one energy input to several outputs in an industry For example a given amount of energy input AH is required to produce essentially two products from a sheep farm wool tonnes and meat tonnes The problem arises when the energy input AH has to be allocated to the different outputs tonnes in order to generate the desired indicators Economicthermodynamic indicators These indicators are hybrid indicators with the energy input being measured in thermodynamic units and the output being measured in terms of market prices That is instead of the output being measured in physical units as for physicalthermodynamic indicators the output is measured in terms of the market value of this output These indicators can be applied to various levels of aggre gation of economic activity product sectoral or national levels EnergyGDP and sectoral energyoutput ratios These energy efficiency measurements of energy input divided by the output can be applied at both the na tional and sectoral levels The energyGDP ratio is the most commonly used aggregate measure of a nations energy efficiency although there has been widespread criticism of the use of this indicator for this purpose The main prob lem with energyGDP as pointed out by Wilson et al 1994 is that it does not measure the underlying technical energy efficiency Other factors such as changes in the sec toral mix in the economy Jenne and Cattell 1983 energy for labour substitution Renshaw 1981 and changes in the energy input mix Liu et al 1992 can influence move ments in the energyGDP ratio and these factors have noth ing to do with technical energy efficiency Recently methods have been developed by Patterson 1993b and others to specifically exclude these extraneous factors from the energyGDP ratio in order to isolate the underlying technical energy efficiency Methodological problems can also emerge in the meas urement of GDP between countries Usually GDP measure ments are commensurated using the exchange rate method which does not necessarily take account of the purchasing 382 What is energy efficiency M G Patterson power of different currencies For this reason it is often ar gued in the literature that the purchasing power parity method of equivalencing GDP should be used to obtain valid crossnational comparisons in the energyGDP ratio Reister 1987 Energy inputoutput ratios are also widely used at the sectoral level and they have exactly the same methodologi cal problems as the energyGDP ratio has at the national level Such sectoral level ratios can be calculated by using official statistics or derived from undertaking algebraic manipulations of inputoutput tables Bullard and Heren deen 1975 These sectoral ratios can be either direct en ergy or total energy ratios Direct energy ratios only take account of the energy directly used by a sector Total energy ratios also take account of the energy indirectly used by a sector ie the energy embodied in the supply of other ma terials and services required by a sector For example a farm will use a certain amount of direct energy to operate its machinery and farm equipment eg diesel to run a trac tor but it will also require other inputs eg fertilizers pes ticides which in turn require energy for their manufacture the energy required to produce these other inputs is called indirect energy Energy productivity ratio This is the reciprocal of the energyGDP ratio ie it is the GDP Y divided by a nations energy consumption E The more goods and services Y an economy produces per unit of energy E the more productive or efficient it is said to be with respect to energy The energy productivity indicator is analogous to the well established labour and capital pro ductivity ratios used in economics and can also be applied at the sectoral level A detailed rationale for monitoring energy productivity changes in the US economy is outlined in a publication by the Joint Economic Committee of the Congress of the United States 1981 The energy productivity ratio is seen as a mechanism for focusing attention on the productive use of energy as a complementary measure to the orthodox capital and labour productivity ratios used in economic analysis The use of the energy productivity ratio in conjunction with labour and capital productivity ratios can provide use ful insights into whether energy inputs act as complements or substitutes to these other factor inputs For example Pat terson 1989 found by using such ratios for New Zealand 196085 that energy and labour inputs acted as mild sub stitutes to each other and energy and capital inputs were mild complements to each other The uncritical use of the energy productivity ratio like that of the energyGDP ratio can lead to misleading con clusions For example the energy productivity ratio may decrease solely because energy is substituting for labour rather than any underlying deterioration in the technical en ergy efficiency To overcome this analytical problem the analyst can calculate the marginal energy productivity ratio by using standard econometric modelling techniques This ratio measures the marginal effect on output by increas ing the energy input AH by one unit 7 Economic indicators The output measurement in the economicthermody namic indicators of energy efficiency is measured in terms of economic value The energy input is still however measured in thermodynamic terms for these hybrid indica tors It could be argued as some economists do that both the input and output measurements should be enumerated in terms of economic value It is argued for example by the Joint Economic Committee of the Congress of the United States 1981 that the energy dollarsGDP ratio is a more accurate reflection of the economic productivity of energy provided that energy prices reflect energy supply and demand forces ie more accurate in comparison to the energy inputGDP ratio It is argued by Turvey and Norbay 1965 and Berndt 1978 that the use of energy prices instead of thermody namic units to measure the energy input provides a solu tion to the energy quality problem ie the problem of validly adding up energy inputs of different qualities This analytical problem is discussed in the next section of this paper in relationship to the fundamental problem it creates in using energy efficiency indicators In brief Turvey and Norbay 1965 and Bern 1978 suggest the use of ideal prices for measuring the energy inputs These ideal price weights reflect either the marginal rates of transformation in production or marginal rates of substitution in consump tion of different energy forms The use of ideal prices to measure energy inputs does however appear to be prob lematical on an operational level due to difficulties in calcu lating these ideal prices in any measurable consistent and assumption free manner There is also evidence that such ideal prices are unstable over time unlike thermodynamic measures of energy which remain constant Beyond the theoretical and operational problems with using prices for measuring energy inputs in efficiency indi cators it could be argued on axiomatic grounds that a pure economic indicator of energy efficiency is not truly an en ergy efficiency indicator Rather it is an economic effi ciency indicator because it is fully enumerated in economic value terms and therefore it should be immediately dis missed as a candidate measure of energy efficiency The most widely advocated pure economic indicator of energy efficiency which has been proposed in the literature is national energy input national output GDP This indicator which is the direct analogue of the energy inputGDP ratio was proposed by the Joint Economic 7The term marginal energy productivity ratio is exactly equivalent to the term marginal product of energy used in economics That is the extra output obtained by employing one extra unit of energy The concept of marginal product can also be applied to other factors of production such as capital or labour What is energy efficiency M G Patterson 383 Committee of the Congress of the United States 1981 al though the Committee was fairly cautious about its widespread use due to the unpalatable assumptions which underpin its use Other pure economic indicators of energy efficiency could be developed at both the national and sec toral levels by simply converting the energy input measure ments to monetary units by using appropriate energy prices These other indicators would be analogous to their phys icalthermodynamic indicator counterparts Although this is possible to the authors knowledge these types of indi cators have not to date been developed for monitoring en ergy efficiency Another possibility suggested by the Joint Committee 1981 is to construct an energy consumer cost savings measure This is seen to have the advantage of directly in forming the public as to how much money has been saved from improvements in energy efficiency It is argued that such an indicator will express the energy efficiency meas ure in terms that everyone can understand money gained from energy efficiency These economic indicators could be developed at the national level andor for particular sectors in the economy Methodological issues in operationalizing energy efficiency indicators There are a number of persistent methodological problems and issues associated with the operationalization of the en ergy efficiency indicators outlined in the previous sections of this paper Most of these methodological problems are common to the full range of energy efficiency indicators and some are just common to a particular type of energy ef ficiency indicator Policy analysts and other practitioners have tended to ignore andor not fully appreciate the impli cations of these methodological problems when attempting to use such energy efficiency indicators Valuation and value judgements The implication in some of the literature is that the thermo dynamic measures of energy efficiency are somehow ob jective and free of value judgements This is true in one sense as given the a priori definition of energy efficiency according to a particular thermodynamic formula two independent observers will obtain the same answer when calculating an efficiency index s This of course assumes that they are both competent at undertaking the calculations and the problem is unambiguously defined Furthermore the thermodynamic efficiency will remain constant over histor ical time and not be subject to changes eg an enthalpic ef ficiency of 20 in 1960 will still be 20 in 1996 This is in contrast to energy efficiency measures that incorporate eco nomic units which change as peoples preferences and tastes change and hence market prices change 8Babbie 1975 refers to this phenomenon of the observers arriving at the same conclusion if the ground rules are agreed upon as intersubjectivity rather than objectivity This is because choice of the ground rules them selves involves subjeclive judgements Nevertheless it is false to assert that thermodynamic measures of energy efficiency are free of human values and perceptions The most common way to define thermody namic energy efficiency in general terms is Useful energy output Energy input Of key importance in considering this ratio is what consti tutes a useful energy output The definition of useful im plicitly requires some assignment of human values in order to define what is considered to be a useful output Socalled unuseful or waste energy eg waste heat does not enter into the calculation of thermodynamic energy efficiency Hence in all thermodynamic energy efficiency definitions there is an implicit value judgement Boulding 1981 succinctly summarizes this issue in his criticism of thermodynamic measures of energy efficiency in social contexts In applying physical concepts like energy to social and eco nomic systems certain pitfalls have to be avoided some of which are very easy to fall into In the first place it is very im portant to recognise that all significant efficiency concepts which are based on purely physical inputs and outputs may not be significant in human terms or at least the significance has to be evaluated The more output per unit of input the more efficient we suppose it to be The significance of the ef ficiency concept however depends on the significance of the outputs and inputs in terms of human valuations Once it is accepted that valuations and value judgements are an integral part of any definition of energy efficiency the next question that can be asked is what is the appropri ate way to assign value to energy inputs and outputs of a particular process It is increasingly being recognized that the value of an energy input should be measured in terms of how much end use service it can deliver eg ECNZ 1992 None of the thermodynamic indicators of energy efficiency measures output in terms of an adequate index of end use service delivery lnstead they measure the value or quality of an energy source in terms of an arbitrarily chosen nu meraire heat content AH a work potential AG or an ideal limit which is defined by the restrictive assumption of infinitely slow processes Obviously neither heat content AH nor work is the only required end use of energy in the economy so therefore a methodology needs to be devel oped to take account of all end uses of energy in the eco nomy eg light sound mechanical drive heating chemical reduction refrigeration pumping and so forth Energy quality problem The energy quality problem is encountered when attempts are made to measure energy efficiency in complex eco nomic systems That is in systems or processes where there are many sources and end uses of energy of differing qual ities Before any energy efficiency calculations can be made these energy forms need to be commensurated or ad justed in terms of energy quality 384 What is energy efficiency M G Patterson This problem always emerges when using enthalpic measurements AH which is the most common way of measuring energy Enthalpic measurements AH only measure the heat content of energy forms and do not nec essarily make any distinction between low quality energy sources such as coal and higher quality energy sources such as electricity From this basis it has consequently been argued that energy when measured in enthalpic terms AD cannot be added up because it has different qualities This problem has variously been called the apples and or anges or aggregation problem Leach 1975 Roberts 1979 The energy quality problem is therefore a fundamental problem in constructing conceptually sound energy effi ciency indicators It is a focus of concern in the construc tion and use of all energy efficiency indicators whether they be at the macrolevel or microlevel At the macrolevel for example the energy quality problem arises in the calculation of the energyGDP ratio when the energy input aggregate is being calculated In this case there are many primary energy inputs into the econ omy of differing qualities Care needs to be taken in aggre gating these primary energy inputs and ensuring that adjustments are made for varying qualities Quite often ana lysts ignore this matter and consequently spurious results are achieved particularly when major shifts in the mix of primary energy inputs into the economy are being analysed If for instance the change in the New Zealand energyGDP ratio is calculated in enthalpic terms it increased only 1545 from 1960 to 1987 but if the energyGDP ratio is calculated taking account of energy quality it increased by 2026 Patterson 1993b The 481 difference between the two figures is quite significant and this discrepancy would not be acceptable when calculating other macro level aggregates such as the Consumer Price Index The energy quality problem is perhaps more acute and problematical at the microlevel where the analyst is at tempting to compare the energy efficiency of several pro cesses with energy inputs of different qualities and possibly with energy outputs also of different qualities For example take the relatively simple case of comparing the energy effi ciencies of three space heating technologies refer to Figure 1 9 1 electricity heat pump space heat 2 electricity resistance heater space heat 3 natural gas enclosed burner space heat In comparing the enthalpic efficiencies of processes 1 and 2 we can validly deduce that process 1 is more efficient than process 2 That is in using electricity to produce space heat the heat pump technology with an enthalpic efficiency of 333 is more efficient than the resistance heater tech 9This example is a simple case as it involves only one type of output space heat The energy quality problem becomes a more complex pro position when you are comparing processes that have different types of outputs eg space heat light motive power In this situation it is even more difficult to compare the relative energy efficiency of all of the pro cesses with each other Ambient heat 73 MJ Heat pump Waste heat 3 MJ Enthalpic efficiency 333 ba first Quality adjusted efficiency 267 first Resistance heater Enthalpic efficiency 100 ba second Quality adjusted efficiency 80 third Waste heat 25 MJ 8 Natural gas 125 MJ Enclosed burner Enthalpic efficiency 80 ba third Quality adjusted efficiency 107 second Figure 1 Ranking of three processes using enthalpic efficiency and quality adjusted measures a aThe following quality coefficients were used in these calculations space heat 080 electricity 100 and natural gas 060 The quality adjusted efficiencies of each process are therefore process I 080 x 100I 00 x 30 267 process 2 080 x 100100 100 80 process 3 080 100060 125 107 nology at 100 This comparison can be validly made be cause we are comparing like with like both processes have the same input electricity and have the same output space heat In fact it does not really matter if the units of electricity or space heat are measured in enthalpic units A or any other measurements as long as the same units are consistently used in measuring both processes For ex ample the electricity units could be measured in terms of kilowatt hours and exactly the same relative efficiencies would result in comparing processes 1 and 2 The energy quality problem however emerges when one attempts to compare the relative efficiency of process 3 with processes 1 and 2 This is because process 3 has nat ural gas as an input not electricity as do processes 1 and 2 This means the analyst is confronted with the problem of comparing two energy inputs of different qualities electri city versus natural gas and the conventional enthalpic measures do not take account of these quality differences Consequently the enthaipic efficiency indicator provides an invalid measure of the comparative energy efficiency of these processes Once energy quality factors have been taken into account the relative order of the energy efficien cies of these technologies changes Instead of natural gas space heat being the least efficient process as measured by the enthalpic indicator it is now the second most effi cient process once energy quality is taken into account A recent paper by Patterson 1993a reviews the differ ent approaches for dealing with the energy quality problem including thermodynamic measures and their modern derivatives OECD thermal equivalents and fossil fuel equivalents Each of these approaches were critically exam ined and found to be inappropriate for measuring energy quality in complex economic systems where a whole variety of processes sources and end uses are concurrently used In Pattersons 1993a paper the quality equivalent methodology was also presented as a candidate method for resolving the energy quality problem in this type of situation Bounda problem Boundary assumptions are implicit in the use of any of the energy efficiency indicators On the output side as was pre viously pointed out only useful energy is included in the calculations On the input side however the situation be comes even more problematical as often quite arbitrary and poorly justified boundaries are drawn First when cal culating energy efficiency indicators only certain energy inputs are considered and others are considered to be out side the studys boundary Noncommercial energy inputs are often excluded ie energy inputs that are not acquired through the market exchange process For example in New Zealand wood energy inputs are often not included in en ergy statistics and hence not included in indicators which use such statistics This is because a significant amount of wood is obtained free of charge from wood processing in dustries scavenged from demolition sites collected from beaches and so forth Solar energy is another energy input often excluded from energy efficiency indicators because it is considered What is energy eiciency M G Patterson 385 to be free This is of course a misconception as there is often a considerable capital investment and hence financial cost in capturing solar energy eg in the use of solar water heaters In addition to minor uses such as solar water heat ing solar energy is also a major input into pastoral horti cultural and forestry industries It is converted via photosynthesis to chemical energy but this energy is ex cluded from official statistics and therefore energy effi ciency indicators because it is considered to be a free source of energy Another dimension of the boundary problem highlighted by the IFIAS 1974 is how far back to trace primary en ergy inputs For example for energy products such as re fined oil do we take account of the energy losses in the refining of the oil if we do take account of these losses in this example then the energy input measurement AH of an energy efficiency indicator will increase This will in turn lead to a decrease in the measured energy efficiency of any process that uses refined oil If such factors are not taken into account by the energy efficiency methodology being employed spurious results could emerge if there is a major shift towards or away from the use of such refined oil products Another more philosophical example of how far back to trace energy inputs is whether to track primary energy inputs back to flows of solar energy inputs For example do we take account of the solar inputs that drive the hydrological cycle to produce hydroelectricity Some analysts such as Costanza 1980 a leading ecological eco nomist suggest we should take account of solar energy in puts in this way It should be noted that these issues of how far back to trace energy inputs are for the most part re solved by using the quality equivalent methodology J0 Joint production problem The partitioning or joint production problem refers to the difficulty of allocating one energy input to several or mul tiple outputs of a process or system This problem is par ticularly encountered in the calculation of physical thermodynamic energy efficiency indicators eg when cal culating the energy input Aoutputkg indicator for an industry that produces multiple outputs For instance a given amount of energy M J is required to produce essen tially two products from a sheep farm wool kg and meat kg The problem arises when the energy input M J has to be allocated to the outputs kg The IFIAS 1974 recom mended four possible conventions for resolving the parti tioning problem IIt is beyond the technical scope of this paper to fully justify this state ment It can however be formally justified by using the mathematics of the quality equivalent methodology This justification hinges on the fact most primary energy inputs are nonbasic energy inputs as defined by Sraffa 1960 and these inputs play no role in determining the quality coef ficients of other energy forms in the reference energy economy Therefore they need not be included in the reference energy economy for determin ing the quality coefficients and consequently the systems boundaries need not encapsulate these inputs 386 What is energy efficiency M G Patterson 1 Assign all energy requirements to the output of interest 2 Assign energy requirements in proportion to financial value or payments 3 Assign energy requirements in proportion to some physical parameter characterizing the system eg weight volume energy content 4 Assign energy requirements in proportion to marginal energy savings which could be made if the good or ser vice was not provided All these conventions are very arbitrary and none of them has gained widespread acceptance Regression analysis has provided a useful tool for over coming this partitioning problem where the inputs or out puts are produced in quantities not proportional to each other For example Cleland et al 1981 used regression analysis to allocate energy inputs to multiple products from food factories Regression analysis has also been used suc cessfully by others Jacobs 1981 Rao et al 1981 in ad dressing the partitioning problem However when the inputs or outputs are proportional or near proportional to each other eg in the case of meat and wool production from a sheep farm the problem is said to be confounded and cannot be solved by regression analysis This type of regression analysis can usually only be applied at the indi vidual factory level using daybyday longitudinal data There usually is not sufficient data available from official statistics either longitudinal or crosssectional to under take such analyses at the sectoral level Technical or gross energy efficiency Most of the indicators of energy efficiency outlined in this paper measure gross energy efficiency of a process system or economic sector As recently pointed out by Wilson et al 1994 in this journal this can lead to difficulties and mis understandings in interpreting these indicators For ex ample indicators of gross energy efficiency such as the energyGDP ratio include a number of other structural fac tors that can significantly affect the numerical magnitude of the indicator but they have nothing to do with the underly ing technical energy efficiency of the economy Policy ana lysts and commentators are often more concerned with the technical improvements in energy efficiency rather than ex traneous structural factors such as sectoral mix changes en ergy input mix changes and energyforlabour substitution processes all of which affect the aggregate measure of en ergy efficiency Liu et al 1992 and Patterson 1993b among others have recently devised methods for isolating this underlying technical energy efficiency For example a study by Patterson and Wadsworth 1993 found that New Zealands energyGDP ratio in creased by 3782 over the 197990 period mainly due to effects other than technical energy efficiency change refer to Figure 2 By far the most influential effects were due to the restructuring of the economy towards more energy in tensive sectors 2672 increase In comparison the dete rioration in technical efficiency technical change residual only contributed to a 69 upward movement in the New Zealand energyGDP ratio Therefore in the New Zealand case the gross energyGDP ratio is highly misleading as an indicator of technical improvements of energy use even though some commentators and politicians use it for this purpose Studies of other countries eg by Wilson et al 1994 Schipper et al 1990 have isolated the technical en ergy efficiency of the energyGDP ratio and unlike the New Zealand situation they have found a consistent improve ment in the underlying technical energy efficiency over the 1970s to 1990s Nevertheless in these countries structural effects have still significantly contributed to changes in the energyGDP ratio eg for the USA Schipper et al 1990 UK Bending et al 1987 and Australia Wilson et al 1994 The same phenomenon occurs with energy efficiency in dicators at both the sectorai and product levels For ex ample the energy intensity MJkg of a factory output may increase because of greater mechanization and hence en ergy use rather than any deterioration in the technical effi ciency of machinery in utilizing energy Similarly a sectoral energyoutput MJUS ratio may also increase due to a movement towards more energy intensive products in that sector Both the technical and gross energy efficiency indicators are equally valid but they are designed to analyse different types of issue For example if the policy analyst is explor ing the broader issues of societal levels of energy use as they relate to resource depletion and sustainability issues a gross energy efficiency indicator eg energyGDP may be more appropriate and should not immediately be dismissed However if one is analysing the efficacy of targeted energy conservation programmes where the focus is quite obvi ously on improving technical levels of energy use then a technical energy efficiency indicator is more appropriate Conclusions Energy efficiency is now a central focus of many national energy policies and at the forefront of the debate on energy sustainability issues but surprisingly little serious attention has been given to defining and measuring the concept If energy efficiency policy objectives are going to be properly set in place and progress towards them monitored theoret ically sound operational definitions of energy efficiency need to be developed This paper has however shown that there are number of critical methodological problems that stand in the way of the establishment of such operational indicators of energy efficiency More attention needs to be given by policy analysts and others to addressing and over coming these methodological problems Thermodynamic indicators of energy efficiency unless they are adjusted for energy quality are very limited at the macrolevel because they do not allow for the ready com parison of energy efficiency across processes which have different energy inputs and outputs Physicalthermody namic indicators whereby the output is measured in phys ical units which reflect the desired end use service of the What is energy efficiency M G Patterson 387 40 30 t q 20 Q 10 Energy input All other Metals Petrochem mix effect sectors I Sectoral mix effect Figure 2 Components of change in New Zealands energyGDP ratio 197990 Household Technical Total residual change residual process are often more useful However these indicators only allow for the comparison of the efficiency of processes which require the same end use service and hence physical thermodynamic indicators are restrictive as general meas ures of energy efficiency Economicthermodynamic indi cators such the energyGDP ratio are more useful for macrolevel policy analysis but often encounter problems with separating the structural effects from the underlying technical energy efficiency trends The energy quality problem is a fundamental problem across all energy efficiency indicators when trying to com pare processes with different quality inputs and outputs In particular the potency of thermodynamic indicators as macrolevel indicators really depends upon the successful resolution of this problem and until this is achieved ther modynamic indicators will remain only useful at the pro cess level of analysis The quality equivalent methodology developed by Patterson I 983 1991 1993a is advocated as an appropriate way of commensurating energy inputs and outputs in terms of their quality This methodology has been specifically designed to measure energy quality in complex economic systems which usually are the context for macrolevel policy studies Other methodological prob lems are less critical to measuring energy efficiency but nevertheless need to be carefully considered by policy ana lysts before attempting to measure energy efficiency at the macrolevel Acknowledgements I would like to thank Professor Donald Cleland Depart ment of Process and Environmental Technology Massey University Associate Professor Gerald Carrington De partment of Physics University of Otago and an anonym ous referee for their valuable comments Any remaining errors of course are the authors responsibility References Ahem J E 1980 The Exergy Method of Energy Systems Analysis John Wiley New York Andresen B Berry R S and Salmon P 1977 Optimisation processes with finitetime thermodynamics in Fazzolare R A and Smith C B eds Energy Use Management Proceedings of the International Con erence Pergamon Press New York Babble E R 1975 The Practice of Social Research Wadsworth Belmont Bejan A 1980 Secondlaw analysis in heat transfer Energy The Inter national Journal 5 8 721732 Bending R C Cattell R K and Eden R J 1987 Energy and structural change in the United Kingdom and Europe Annual Review of Energy 12 185222 Berndt E R 1978 Aggregate energy efficiency and productivity meas urement Annual Review of Energy 3 225249 Berry R S and Fels M F 1973 Energy cost of automobiles Science and Public ktirs December I I0 Boulding K E 1981 Evolutionary Economics Sage Publications CA Bullard C W and Herendeen R A 1975 The energy cost of goods and services an inputoutput analysis for the USA 1963 and 1967 Energy Policy 3 4 268 78 Cleland A C Earle M D and Boag 1 F 1981 Application of linear re gression to analysis of data from factory energy surveys Journal q Food Technology 16 481 492 388 What is energy efficiency M G Patterson Collins C 1992 Transport Energy Management Policies Potential in New Zealand Ministry of Commerce Wellington Costanza R 1980 Embodied energy and economic valuation Science 210 12191224 Electricity Corporation of New Zealand 1992 The Developing Market br Energy Efficiency in New Zealand Electricity Corporation of New Zealand Wellington Groscurth H M KOmmel R and van Gool W 1989 Thermodynamic limits to energy optimisation Energy The International Journal 14 2 241258 Gyfiopoulos E P Lazaridis L J and Widmer T F F 1974 Potential Fuel Effectiveness in Industry A Report to the Ford Foundation Energy Pol icy Project Ballinger Publishing Company San Francisco Harris G Gale S Allan R and Lucas M 1993 Promoting a Market for Energy Efficiency Report to the Oficials Committee on Energy Policy Ministry of Commerce Wellington Horsley M 1993 Engineering Thermodynamics Chapman and Hall London International Federation of Institutes for Advanced Study I 974 Energy Analysis Workshop on Methodology and Conventions Report No 6 IFIAS Stockholm Jacobs P W 1981 Forecasting energy requirements Chemical Engin eering 80 6 9799 Jaynes E T 1989 Clearing up mysteries the original goal in Skilling J ed Maximum Entropy and Bayesian Methods Klumer Academic Pub lishers Boston MA Jenne C A and Cattell R K 1983 Structural change and energy effi ciency in industry Energy Economics 5 2 114123 Joint Economic Committee of the Congress of the United States 1981 A National lndex br Energy Productivity US Government Washington DC Kay R L A and Scholenhls R J 1980 The second law efficiency of a heat pump system Energy The International Journal 5 8 853863 Leach G 1975 Net energy analysis is it any use Energy Policy 3 4 332344 Liu X Q Ang B W and Ong H L 1992 lnterfuel substitution and de composition of changes in industrial energy consumption Energy The International Journal 17 7 68996 Lovins A B 1977 Soft Energy Paths Towards a Durable Peace Harper Colophon Sydney Often R J 1978 An improved definition of energy efficiency in Energy Conservation Source Book Ministry of Energy Wellington Patterson M G 1983 Estimation of the quality of energy sources and uses Energy Policy 11 4 346359 Patterson M G 1989 Energy Productivity and Economic Growth An Analysis of New Zealand and Overseas Trends Ministry of Energy Wellington Patterson M G 199 L A Systems Approach to Energy Quality and Effi cient T Unpublished revision of a PhD thesis Victoria University of Wellington Patterson M G 1993a Approaches to energy quality in energy analysis International Journal of Global Energy Issues 5 I 1928 Patterson M G 1993b An accounting framework for decomposing the energytoGDP ratio into its structural components of change Energy The International Journal 18 7 741761 Patterson M G and Wadsworth C 1993 Updating New Zealands En ergy Intensity Trends What has Happened Since 1984 and Why En ergy Efficiency and Conservation Authority Wellington Rao M A Goel U K Vergara W Jordan W K and Cooley H J 1981 Direct energy consumption for processing of food products via multiple regression in Agricultural Energy Vol 3 Food Processing American Society of Agricultural Engineers St Josephs MI Reister D B 1987 The link between energy and GDP in developing countries Energy The International Journal 12 6 427433 Renshaw E F 0981 Energy efficiency and the slump in labour produc tivity in the USA Energy Economics 3 1 3642 Roberts W N T 1979 Overall energy balances and the addingup problem in Workshop on Energy Data of Developing Countries Interna tional Energy Agency Paris Schipper L Howarth R B and Geller H 1990 United States energy use from 1973 to 1987 the impact of improved efficiency Annual Review of Energy 15 445604 Schurr S H 1984 Energy use technological change and productive effi ciency an economichistorical interpretation Annual Review of Energy 9 40945 I Sioshansi F 1986 Energy electricity and the US economy emerging trends The Energy Journal 7 4 8149 Slesser M 1982 Dictionary of Energy Schocken Books New York Sraffa P 1960 Production of Commodities by Means of Commodities Cambridge University Press Cambridge Sussman M V 1977 Availability analysis in Fazzolare R A and Smith C B eds Energy Use Management Proceedings of the International Conference Pergamon Press New York Terry S 1991 Making a Market for Energy Efficiency New Zealand Planning Council Wellington Turvey R and Norbay A R 1965 On measuring energy consumption Economic Journal 75 787793 van Gool W 1980 Thermodynamic aspects of energy conversion En ergy The International Journal 5 8 783792 Wilson B Trieu L H and Bowen B 1994 Energy efficiency trends in Australia Energy Policy 22 4 287295 Wu C 1988 Power optimisation of a finitetime Camot heat engine En ergy The International Journal 13 9 681687 Appendix Brief explanation of the quality equivalent methodology The purpose of the quality equivalent methodology I is to define an energy unit which allows energy inputs and outputs to be com pared on a common basis This energy unit is called a quality equivalent and is defined by solving a system of simultaneous lin ear equations These equations which are termed a reference sys tem quantify the flow of energy in national energy systems eg the 1995 UK energy system As such there is a description of the flow of energy from primary energy sources to delivered energy and eventually to end uses of energy t Fuller explanations of the quality equivalent methodology are contained in Patterson 1993a and also in an earlier publication in Energy Policy by Patterson 1983 Reference system equations The flow of energy in any complex system such as a national en ergy system can be quantified by a system of simultaneous linear equations represented by Xle0 where X matrix m n of m processes describing the conversion of energy between n types of energy The energy flows are measured in AH terms with inputs entered as nega tive entries and outputs as positive entries e column vector n 1 of quality coefficients of each en ergy type The quality coefficients are measured in terms of EAHunits and are determined by solving the simulta neous equations residual vector m 1 The residual expressed in quality equivalents E for each process For a process with an ef ficiency equalling the systems average e 0 for a pro cess efficiency less than the systems average e 0 and for a process efficiency greater than the systems average e0 This system of simultaneous equations needs to be solved so as to determine the quality coefficients for each of the energy types ie to obtain a solution vector I This presents a number of problems First the system of equations is nearly always overdetermined as there are more conversion processes m than energy types n Therefore deterministic solution methods such as those used in Leontiefstyle input output analysis are not suitable solution methods Second the system of equations are homogeneous as the right hand side of the equations is a vector of zero entries For this reason the trivial solution of 1 0 is always a possible solu tion but not meaningful The key to solving the equation is to avoid the trivial solution by setting one of the quality coefficients to unity and transferring the resultant vector to the other side of the system of equations 12 The solved quality coefficients are expressed in terms of multiples of the variable which has been transferred to the righthand side These multiples are called qual ity equivalents Any one of the specific coefficients in the refer ence system can be used as the quality equivalent unit For a properly specified system of equations it does not matter which coefficient is set to unity as the relativities between the quality co efficients remain constant Quality equivalent unit and quality coefficients 13 The concept of the quality equivalent unit is pivotal in the QEM The quality equivalent unit is the measuring rod which allows energy forms to be compared on a common basis in terms of their energy quality Energy inputs and outputs have been traditionally measured in terms of their heat content AH which takes no ac count of energy quality To convert energy inputs and outputs measured in heat units AH to quality equivalent units E they need to be multiplied by the quality coefficients EAll obtained from solving the above specified system of equations In general the quality coefficients EAH provide a measure ment of the quality of energy inputs and outputs The specific meaning that can be attached to the numerical value of each qual ity coefficient depends on the type of energy inputoutput For primary energy inputs the quality coefficient EoutAHin is the relative efficiency at which a primary energy input AHin is con verted to energy enduses Eout in the reference system The higher the energy quality of a primary energy input the more end use energy it will produce For example a primary energy input such as natural gas is usually more efficient or productive at pro 12The most straightforward solution method although not the most reli able is to solve the equations by using least squares regression In this method proposed by Patterson 1983 each coefficient is in turn set to unity to generate different regression models Out of all of these regres sions the model with the highest R 2 is selected for final use Other more reliable solution methods have been developed by Patterson 1991 13These concepts have direct analogues in economic thinking quality equivalent E monetary value quality coefficient Eunit of energy relative price Sunit of commodity What is energy efficiency M G Patterson 389 ducing end uses of energy than lower quality energy inputs such as coal That is one unit of natural gas AHin will produce more end use energy Eout than one unit of coal AHin Therefore nat ural gas will have a higher quality coefficient EoutJin than that for coal For an end use of energy which does not feedback into the system its quality coefficient EinAHout is the total embodied energy required to produce that end use For example a typical high quality end use such as light energy requires a greater input of direct and indirect energy AHin to produce one useful output of energy Eout The QEM provides for an integration of the concepts of qual ity of inputs and quality of outputs within one framework In fact it is argued that it is impossible to rigorously measure the quality of either an input or output without reference to each other Again analogies can be drawn with economic thinking with re spect to how equilibrium prices enable supplyside cost and de mandside utility ideas to be reconciled A simple numerical example Consider the notional reference system of energy conversions por trayed by Figure 3 Algebraic equations can be used to describe the conversion of inputs A to outputs AH of energy for each process in the reference system In the following equations the in puts are arranged on the lefthand side and the output on the right hand side with feedbacks of energy required by each process denoted by underlining I bl1450b70lObsO20e I b41350 2 b6600 b8002 e 2 b4200 3 b5200 b7080 b8001 e 3 b4050 4 b 21600 b8001 e 4 b 61400 5 b 312500 b7020 e 5 b 510000 6 b4600 b8004 e 6 b7600 7 b6400 b8003 e 7 b7300 8 b5800 bs00 e 8 b7480 9 b6400 b7004 e 9 bs060 10 b58000 b7004 el0 b8800 11 b41000 b8004 ett b9100 This system of simultaneous equations can be solved and ex pressed in terms of multiples of any of the energy forms in this particular case delivered electricity equivalents 1 b I 08823 hydroelectricity 2 b 2 03755 wellstream gas 3 b 3 02509 crude oil 4 b 4 10000 delivered electricity 5 b 5 03152 oil products 6 b 6 04314 delivered gas 7 b 7 07813 heat 8 b s 31403 transport 9 b 9 101256 lighting Enduse matching and process efflciencies It becomes evident from solving the equations that not all pro cesses are equally efficient as demonstrated by the existence of nonzero residuals e o The relative efficiency of each process can be calculated by dividing the outputs Eou t by the in puts Ein of each process see Table 1 390 What is energy elciency M G Patterson Primary energy Consumer energy 4 End use energy 7 6 10 Figure 3 Reference system energy conversion processes a aOnly direct energy conversion processes are depicted All conversion processes also require feedbacks of end use energy for their operation Table I Process efficlencles and residuals for the simple numerical example Process Input Process output Relative efficiency Residual Hydroelectricity Delivered electricity Oj 10000 e t 0 Delivered Gas Delivered electricity 2 07544 e 2 06512 Oil products Delivered electricity O 3 03885 e 3 07869 Wellstream gas Delivered gas 4 10000 e 4 0 Crude oil Oil products 05 10000 e 5 0 Delivered electricity Heat 06 07652 e 6 14381 Delivered gas Heat 7 12879 e 7 05239 Oil products Heat 08 13224 es 09142 Delivered gas Transport 9 10725 e 9 01273 Oil products Transport Oi0 09950 el0 01273 Delivered electricity Lighting Oil 10000 ell 0 Processes that have relative efficiencies of greater than one 0 1 are more efficient than the systems average and those that have relative efficiencies less than one 0 1 are less effi cient than the systems average By using these relative efficien cies it is possible to rigorously match end uses and sources of energy in accordance with the type of ideas promoted by Lovins 1977 For example the most efficient way of providing heat is by using oil products 0 8 13224 whereas the least efficient way of providing heat is by using electricity 0 6 07652 O artigo de MG Patterson oferece uma análise aprofundada sobre o conceito de eficiência energética destacando sua relevância para políticas energéticas nacionais e para a sustentabilidade energética em geral O autor inicia formulando a distinção entre eficiência técnica e eficiência bruta de energia salientando que indicadores normalmente utilizados como a relação 𝐸𝑛𝑒𝑟𝑔𝑖𝑎 𝑃𝐼𝐵 Podem ser suscetíveis a distorções devido a fatores estruturais e não necessariamente refletir melhorias na eficiência técnica subjacente Quanto aos indicadores termodinâmicos de eficiência energética apesar de inicialmente parecer uma medida lógica apresentam limitações significativas Por exemplo a eficiência da primeira lei que se baseia no conteúdo de calor da entrada e saída do processo não leva em conta a qualidade da energia tratando diferentes fontes de energia de forma igual Já a eficiência de segunda lei tenta contornar esse problema considerando o trabalho máximo realizável em um processo No entanto essa abordagem também possui limitações como a incapacidade de considerar entradas de energia indiretas Já em relação aos indicadores físicotermodinâmicos que tem como objetivo medir a eficiência energética em termos de unidades físicas dos produtos finais refletindo assim o serviço final demandado pelos consumidores Esses indicadores são considerados mais adequados que os indicadores termodinâmicos tradicionais pois capturam diretamente o que os consumidores valorizam em termos de serviço final ao invés de medir apenas o conteúdo de calor ou potencial de trabalho Por exemplo no transporte de carga a eficiência energética pode ser medida em termos de energia por tonelada transportada por quilômetro Essas medidas físicas podem ser objetivamente comparadas ao longo do tempo sem as dificuldades encontradas ao se utilizar indicadores econômicos de eficiência energética No entanto a aplicação desses indicadores híbridos deve ser feita de forma setorial levando em conta as diferentes especificidades de cada setor Sobre os indicadores econômicos temos que eles medem a saída em termos de valor econômico enquanto a entrada de energia ainda é medida em termos termodinâmicos Alguns economistas defendem a medição tanto da entrada quanto da saída em termos de valor econômico argumentando que isso oferece uma solução para o problema da qualidade da energia No entanto há desafios operacionais e teóricos nessa abordagem incluindo a instabilidade dos preços ideais de energia ao longo do tempo Além disso alguns questionam se indicadores puramente econômicos realmente refletem a eficiência energética ou apenas a eficiência econômica Apesar das discussões teóricas ainda não foram desenvolvidos indicadores econômicos puros de eficiência energética para monitoramento Outra proposta é a criação de um indicador de economia de custos do consumidor de energia que informaria diretamente ao público quanto dinheiro foi economizado com melhorias na eficiência energética Um ponto crucial abordado é a questão metodológica associada à medição da eficiência energética O autor discute a complexidade na definição de indicadores precisos dada a diversidade de abordagens conceituais e operacionais Destaca o uso da análise de regressão como uma ferramenta útil para a alocação de energia em produtos e processos Entretanto reconhecese as limitações dessa abordagem especialmente quando os insumos e resultados energéticos não mantêm uma relação diretamente proporcional Uma contribuição significativa do artigo é a introdução da Metodologia de Equivalência de Qualidade QEM uma abordagem que permite a comparação de diferentes formas de energia levando em conta sua qualidade e eficiência Patterson exemplifica a aplicação da QEM em um contexto prático demonstrando como ela pode ser utilizada para avaliar a eficiência energética de diferentes processos e setores fornecendo insights valiosos para políticas energéticas e estratégias de eficiência A conclusão do artigo enfatiza a importância de desenvolver definições e indicadores operacionais teoricamente sólidos para a eficiência energética dada sua centralidade nas políticas energéticas nacionais e nos debates sobre sustentabilidade energética O autor destaca as dificuldades metodológicas existentes nesse processo e aponta a necessidade de superálas para estabelecer indicadores eficazes de eficiência energética Além disso ressaltase a distinção entre eficiência energética técnica e bruta destacando que a última pode ser influenciada por fatores estruturais extrínsecos o que pode levar a interpretações equivocadas dos indicadores Ao abordar as dificuldades específicas na mensuração da eficiência energética como a questão da qualidade da energia é sugerido a metodologia de equivalência de qualidade como uma abordagem adequada para mensurar insumos e resultados energéticos em termos de sua qualidade O artigo de MG Patterson oferece uma análise aprofundada sobre o conceito de eficiência energética destacando sua relevância para políticas energéticas nacionais e para a sustentabilidade energética em geral O autor inicia formulando a distinção entre eficiência técnica e eficiência bruta de energia salientando que indicadores normalmente utilizados como a relação 𝐸𝑛𝑒𝑟𝑔𝑖𝑎 𝑃𝐼𝐵 Podem ser suscetíveis a distorções devido a fatores estruturais e não necessariamente refletir melhorias na eficiência técnica subjacente Um ponto crucial abordado é a questão metodológica associada à medição da eficiência energética O autor discute a complexidade na definição de indicadores precisos dada a diversidade de abordagens conceituais e operacionais Destaca o uso da análise de regressão como uma ferramenta útil para a alocação de energia em produtos e processos Entretanto reconhecese as limitações dessa abordagem especialmente quando os insumos e resultados energéticos não mantêm uma relação diretamente proporcional Uma contribuição significativa do artigo é a introdução da Metodologia de Equivalência de Qualidade QEM uma abordagem que permite a comparação de diferentes formas de energia levando em conta sua qualidade e eficiência Patterson exemplifica a aplicação da QEM em um contexto prático demonstrando como ela pode ser utilizada para avaliar a eficiência energética de diferentes processos e setores fornecendo insights valiosos para políticas energéticas e estratégias de eficiência A conclusão do artigo enfatiza a importância de desenvolver definições e indicadores operacionais teoricamente sólidos para a eficiência energética dada sua centralidade nas políticas energéticas nacionais e nos debates sobre sustentabilidade energética O autor destaca as dificuldades metodológicas existentes nesse processo e aponta a necessidade de superálas para estabelecer indicadores eficazes de eficiência energética Além disso ressaltase a distinção entre eficiência energética técnica e bruta destacando que a última pode ser influenciada por fatores estruturais extrínsecos o que pode levar a interpretações equivocadas dos indicadores Ao abordar as dificuldades específicas na mensuração da eficiência energética como a questão da qualidade da energia é sugerido a metodologia de equivalência de qualidade como uma abordagem adequada para mensurar insumos e resultados energéticos em termos de sua qualidade