“TECHNOLOGY FORECASTING – CRISIS ANALYSIS”
Technology Futures & Business Strategy 1st Assessment Project
Michel Godet indicated that qualitative parameters were important in accurate forecasting. Using the available information in the international literature and between 1000 and 1500 words:
1. Mention the qualitative parameters that may be considered in future energy price scenarios. For this purpose take the year 2020 and list, with a brief explanation, the parameters you consider should be included.
2. Which of these parameters can you reasonably quantify? (Attempt to identify at least five parameters)
3. Do you agree with this specific aspect of Godet’s proposition? Why or why not?
4. Evaluate a crisis impact of the accuracy of technology forecasting. Identify the parameters characterizing the crisis aspects. Accordingly, present your opinion about the validity of the forecasts.
5. Using the installed nuclear power data between 1967 and 1987 estimate (using extrapolation techniques) the expected nuclear power time evolution between 1987 and 2007. Comment on the accuracy of your forecasts in relation with the real data. Can you mention any lead time between the major accident of Chernobyl and the reaction of the international electrical power market?
The OPEC oil price rise in 1973 had an important effect on energy use and energy efficiency, although much of the impact was short-lived. In 2003-4 the oil price effectively doubled, reaching $50/barrel for a period and lately it has reached over $90/barrel. A major player now is Gazprom in Russia News has broken that Gazprom will cut supplies of natural gas to Europe unless it is allowed to raise prices by 200% for export customers, (Customers in Russia historically pay much lower prices). Using the available information in the international literature and between 2000 and 2500 words:
6. Describe your measured response to this, as either an energy Supplier or major energy user.
7. Would you say that your response was based upon “out of the box” solutions, or a more conservative, incremental approach?
8. Discuss the relative merits and limitations of each of these possible responses, identifying what you believe the two approaches mean.
9. How this crisis shall influence the future of European economies? How could these effects been mitigated? Make your own forecasts.
Your answers to 6-8 above are based upon assumed positions within organisations which may employ many people. The next part of this question relates to the impact rising energy prices and, perhaps more importantly, the effect of climate change, may have on your own style of living.
10. At a personal/domestic level, can you foresee a situation in which we may consider that for the benefit of all, we may need to make do with less, in terms of capital goods, travel, and perceived acceptable levels of comfort?
Technology Futures & Business Strategy 1st Assessment Project
Based on the Prospective approach and the scenarios method (Godet, 1982), Michel Godet noted the limitations of the classical forecasting concerned with quantification and models (see also Appendix, Table Ap.I). According to Godet, models that only consider quantified parameters do not take into account the development of new relationships and the possible changes in trends. The impossibility of forecasting the future as a function solely of past data is directly related to the omission of qualitative and non-quantifiable parameters such as the wishes and behaviour of relevant actors (Godet, 1982). Furthermore, to structure future scenarios, the variables related to the phenomenon under investigation and the variables configuring its environment should be recognized and analyzed in detail. Besides, the interrelationship among variables, the relative power and fundamental actors, their strategies and available resources as well as the objectives and constraints that must be overcome, should also be taken into account.
By granting energy as a commodity under the view of conventional economic theories, markets and price mechanisms are used in order to allocate the respective resources. More specifically, it is the interaction of demand and supply in the markets that allocates resources and largely shapes prices, and it is the broader ecosystem boundaries that these market interactions take place in. Energy pricing, with energy being perceived either as an input or as a potentially polluting source of our ecosystem, clearly stands upon both the sub disciplines of resource and environmental economics (Sweeney, 2004), also depending on the social, political and technological status of the time being and the time to come until 2020. In this context, one may acknowledge a bundle of parameters that may be considered for configuring the respective future energy price scenarios. What is important to note is that similar to the beliefs of Godet, the parameters involved should be studied in terms of interrelationship, while qualitative and non quantified parameters should be taken into account as well.
As already mentioned, the configuration of prices within a market -the energy market currently discussed- is largely dependent on the supply and demand balance. This is measured by the respective supply and demand tension expressing the status of a commodity in market terms and providing indications concerning potential energy price changes. While high tensions imply prices’ imbalance, the opposite is valid for low tension rates. Hence, in order to evaluate future energy prices on the basis of parameters, one should identify the parameters that influence the supply-demand balance in every of the fields previously acknowledged (i.e. social, political, environmental, economical and technological). In this context, in 1.1 the most influential of the parameters configuring energy prices may be encountered.
Energy markets are largely influenced by the economic growth factors expressed on the basis of Gross Domestic Product (GDP), inflation, interest and unemployment rates. Given the economic growth along with the parameter of demographics (regarding both the population increase and migrations) one may picture the corresponding trend in energy consumption (i.e. the demand side). Following, policy decisions concerning the determination of fuel mix are determinative as far as energy pricing is considered. For instance, fossil fuels continuing to dominate will stimulate stricter pollution prevention legislation measures (e.g. taxation) and policies for tackling climate change and global warming that will raise energy prices.
In parallel, the reinforcement of the respective market holders, potentially leading to strong monopolies, should also be expected. Turning to renewable energy sources may on the one hand -for some of the technologies- imply an adjustment period in order for the corresponding markets to balance and on the other entail significant environmental benefits, in monetary terms as well. Global warming and climate change effects being evident supports the implementation of mitigation measures towards the reduction of greenhouse gas (GHG) emissions, this holding a key role in respect of the future.
Reserves holding a key role in the future configuration of energy prices not only in terms of scarcity, but also in terms of production costs is directly related with the technological development concerning the exploitation of new deposits and the promotion of substitutes. As already implied, the power of existing markets is another key factor while the efficiency and absorption of energy investments -the investment shares and outcomes of research and development efforts should be underlined- must be also taken into account. The factors concerned with the quality of life suggest an additional parameter that may affect energy consumption patterns and one that cannot be easily captured despite of the indices recommended so far (Allen, 1991).
Moreover, as properly put in the Annual Energy Outlook of 2007 (EIA, 2007a), energy markets projections are subject to much uncertainty (unanticipated events). Many of the events that shape energy markets and therefore the price of energy as well cannot be foreseen. These include unexpected weather events and natural disasters (Rezek and Blair, 2008), major innovations and technological breakthroughs (Marbán and Valdés-Solís, 2007; Varandas, 2008), disruptions and whirls in the political level (Stern, 2006) with analogous societal consequences, the outbreak of a war (Tahmassebi, 1986; Fernandez, 2008) or a nuclear accident, all of them either smouldering or implying blind spots that cannot be directly projected and consequently quantified. Besides, another area of uncertainty is concerned with the fact that even the established trends’ steady evolution cannot be guaranteed.
Summarizing, a brief explanation was presently given on how each of the parameters acknowledged may influence energy pricing. Additionally, an effort was also carried out in order to give a short description of the interrelationship among parameters, this supporting one of Godet’s arguments. Insisting on the interrelationship of variables, several of the parameters previously encountered should be diffused to every major regional energy market, the latter being largely influenced by the relationship between fuel types and energy sectors (see also 1.2).
Eventually, one may result in a rather complex system that encounters the evolution of influential parameters inside the balance between energy types and energy sectors, this revealing the crucial role of energy fuel mix previously discussed. Following, an effort is carried out in order to reasonably quantify some of the parameters acknowledged.
Given the bundle of parameters that are thought to influence future energy pricing, a certain number of them can be quantified. For instance, the parameters of population, economic growth, energy consumption, greenhouse gas emissions, energy reserves, and energy fuel mix can be expressed in numerical terms.
Demographic growth examines how regional and global demography changes over time. According to the United Nations projections (UN, 2006), world population will increase by over 1 billion people in the years to come until 2020, this suggesting an annual increase rate of 1.1%. While in some areas there is a negative population growth to be considered (e.g. European countries), the opposite may be encountered for some of the Asian countries (e.g. India) where overpopulation is met (see for example 2.1 with EIA forecasts). Besides, the migration of people comprises an additional factor influencing energy patterns via the imposition of unequal population distribution already encountered due to birth and mortality rates.
Based on the energy consumption trends ( 2.2), it is expected that energy demand related to all energy products will increase in the years to come, even in such levels that supply may not be able to adequately respond (Asif and Muneer, 2007). In fact, the annual world energy consumption growth is approximately 2% with projections supporting future average rates of 1.1% per annum (EIA, 2007b). In fact, by considering the two of parameters so far examined one may result in the most substantial energy per cap index, clearly establishing the differentiation in energy consumption patterns among world regions (see also question 10).
Furthermore, according to the WEO claims (WEC, 2007) that energy generated from fossil fuels will remain the main energy source (expected to cover almost 83% of global energy demand in 2030) and given the 2020 time horizon, much depends on the appearing constraints of world energy reserves, especially those regarding oil and natural gas. While certain studies sound relieving (WCI, 2007), others questioning the extent of increase in the production outputs ring the alarm of forthcoming peaks within the next one or two decades (Bentley, 2002).
If the latter is valid, the corresponding demand will not be met, prices will rise, inflation, and international tension will become very likely to occur, and inevitably energy users will have to ration (Wirl, 2008). Overall, what the combination of energy mix with energy reserves provides is the measuring of security of supply, the latter configuring the supply and demand tensions, largely shaping energy prices. Besides, targets set in respect of renewable energy sources further penetration also provide a quantification view; e.g. the EWEA target for 22% coverage of the European electricity consumption by 2030 (EWEA, 2006).
Next, expressing economic growth on the basis of gross domestic product (GDP) suggests a constant increase of the former within the range of an average 3% to 4% per year (IMF, 2004), noted during the period from 1970 to 2003. Again, inequity that is to be considered among different world regions is directly related with the previous parameters, illustrating the energy requirements’ variation. A characteristic example considers China demonstrating an average annual percent change of GDP 2.4% greater than the world average. In 2.3, the respective trends of GDP growth up to the year 2020 may be obtained.
Finally, the environmental impact of energy use being expressed on the basis of GHG emissions not only considers the fuel mix and energy consumption but also takes into account the technology used for energy generation. Taking CO2, an increase of 17Gt in a 34 years period, i.e. from 1970 to 2004 (IPCC, 2007), indicates the strong increasing trend, also presented in 2.4. Given also some of the commitments adopted in order to mitigate the greenhouse effect however (e.g. the Kyoto protocol), further quantification, not relying solely on past trends, is possible. The stimulation of additional mitigation measures until 2020 is rather likely, this both imposing the need for shifting to non-fossil fuels and developing cleaner energy generation technologies.
Considering the various parameters’ trends illustrated above, one may sense that the tensions between supply and demand, this comprising the main driver for energy prices, are going to rise. Energy consumption, GDP and population rates on one hand demonstrate the demand side, while declining reserves and mitigation measures describe the opposite supply side. In between, the decisions for future energy fuel mix patterns, although able to completely reverse the energy market’s status quo, are not thought to radically vary within the next 10 to 15 years. Hence, unless some major changes occur, the rising tensions between supply and demand imply both instability and increase of prices on a global level with strong differentiation to be encountered among different world regions. As far as the degree of energy price variation is concerned, the implementation of forecasting may both incorporate all of the pre-mentioned parameters and provide various scenarios considering each one’s expected future time evolution.
As previously seen, several parameters were acknowledged in order to form future energy price scenarios. While some of them were possible to quantify, others although not quantified were equally important inputs to keep in mind. Apart from the given inaccuracy of data (either high or low levelled) coupled with unstable models and the pertinacity of explaining the future in terms of the past, Godet emphasizes on the lack of a global and qualitative approach concerned with forecasting (Godet, 1982). Although quantitative methods may prove to be reliable enough and reasonably accurate for short term forecasts, the same is not valid for forecasts concerned with longer periods. The greater the distance from the reference point, the more obvious is the inability of quantitative data to provide valid forecasts (see also 3.1).
In this context, it is critical to comment on the relativity of time scales noted among the study of various phenomena. Hence, what may seem short termed for one phenomenon studied may actually comprise a long period forecast for another that appears to be rapidly changing over time. Any case given, the chances of significant changes regarding the environment in which the phenomenon under study develops are considerably higher as the time horizon becomes longer and it would be more or less naïve to solely depend on forecasting methods like the extrapolation of trends.
Furthermore, the complexity of phenomena studied and the interdependence among the influencing parameters calls for the inclusion of both quantitative and qualitative parameters with Godet clearly addressing the complementarity between the prospective and classical forecasting (Godet, 1982). It was in fact during the first section of this part that the analysis of energy pricing configuration revealed the importance of interaction between quantitative and qualitative parameters. Energy price could not be disengaged from the parallel evolvement of parameters that even though not easily quantified, do structure the phenomenon environment (e.g. political, technological, economic, social, legal and other aspects). What must be outlined here is that similar to the scaling of decision making (strategic-long term, innovative-medium term, operational-short term), the role of quantitative data is gradually fading out as we tend to conceptualize the entire phenomenon environment. Hence the broader the view, again the more obvious is the inability of quantitative data to support a reliable forecasting (see also 2.1).
Although in its extreme point of view, Godet’s proposition perfectly fits the ability of diagnosing forthcoming crises. Already extremely difficult to predict a crisis, omitting parameters such as the wishes of relevant actors and other influential factors that cannot be quantified makes it impossible even to sense it. It is in this context that one should not disregard the importance of other forecasting resources -apart from data- including assumptions, insight and judgment, all of them involving the subjectivity factor. If managing to get over the reef of the NIH syndrome, creativity and broad minded thinking are also essential elements for good forecasting.
1973 may be granted as the most pivotal year in energy history. The energy crisis defining the period began on October 17, 1973, when the Arab members of OPEC along with Egypt and Syria, all together comprising OAPEC, decided to place an embargo on shipments of crude oil to nations that had supported Israel in its conflict with Syria and Egypt, mainly targeting at the United States and Netherlands. The result of this decision also brought about major oil price increases. Because of the fact that OPEC was the dominant oil distributor at the time, the price increase implied serious impacts on the national economies of the targeted countries, therefore suggesting an international range crisis. Although the embargo was lifted in March 1974, the effects of the energy crisis, mainly in terms of price increase, lingered on throughout the 1970s, with the Iranian crisis aggravating the situation (see also 4.1).
Suggesting a crisis that was mainly expressed on the basis of high energy pricing, the outcome of the previous questions concerned with the determination of energy price influential parameters may be illustrated. In fact, the impact of a more or less unanticipated event changed the correlation patterns between supply and demand and imposed the attachment of high tensions in the market balance, the latter entailing the high volatility of oil price and its potential outburst ever since (Regnier, 2007). The market structures, the dominance of OPEC and the political tension, all suggest aspects of the crisis illustrating the importance of considering qualitative parameters as well. As Godet well pointed out, one cannot neglect the wishes and decisions of major actors when configuring the future (e.g. OPEC members).
Similar to the 1973 oil crisis, the California energy crisis occurring some 27 years later also revealed the strength of key actors in completely changing what was meant to follow a past trend or ameliorate a past situation. The deregulation of the electricity market in California (during 1998) targeting to decrease the highest of retail prices among the States turned into a complete fiasco that abetted the manipulation of the market by the energy companies. The crisis main characteristics involved very high wholesale prices, interrupted service of customers (rolling blackouts), bankrupt utilities and huge state expenditures, while the crisis main causes were:
- The lack of new generating capacity inside California (California was heavily dependent on energy imports from nearby states (CEC, 2007a)).
- The coincidence of a dry year and natural gas spikes with other market oriented factors (California was largely based on hydro and natural gas for the consumers’ electrification).
- The market structure allowing generators to manipulate wholesale prices in the power exchange market through escalating power plants’ outages that caused market disorder (on the other hand there was a retail price cap that did not allow the investor owned utilities to pass the increasing cost of wholesale purchases to consumers).
- The delay and inability of the regulators to predict the crisis and respond to it (it was only after a certain time that a wholesale cap was set by the Federal Energy Regulatory Commission and an increase of retail prices was allowed to the investor owned utilities).
Emphasizing on the manipulation of the market by the energy generators, in 4.2 one may observe the out of schedule power plant outages rapid increase during the period of the crisis, even exceeding 10GW (approximately 20% of the total installed capacity), responsible for three series of rolling blackouts. No prediction could have captured the 300% and 400% increases of power plants outages. The analogous increase in wholesale prices being the result of the appearing power deficit caused the major suppliers (3 major investor utilities (IOUs)) to be trapped between remarkable wholesale price increases and a fixed retail price (see 4.3).
Further, as seen in 4.3, in the early days of deregulation a relatively smooth trend was to be encountered as far as the wholesale market prices are concerned, this also not implying the rapid increase of prices following. Accordingly, although not influenced to the same extent that the IOUs were, the instant impact to the final consumers must also be considered. Note that according to the rough forecast of retail electricity prices -being based on the respective past data- the increase of retail prices was not to be expected either because deregulation promised for a lowering of prices or because the trend applied entailed much lower prices then the ones actually presented at the time (see also 4.4).
Similar to this, predictions involving oil pricing before the 1973 crisis and relying on extrapolation techniques (Anon, 1973) expected that world energy consumption would keep up in the increasing rates of 5% up to 2000. If having managed to somehow foresee the 1973 oil price increase, the predictions made would not be exclusively based on the past data trend that would undoubtedly provide a misjudgement of future prices (see also 4.5). What actually followed for the years to come (1980 to 2000) was a 20 years mean annual increase rate of 1.7%. Furthermore, if only having used quantitative data, none could have predicted before the crisis that USA would cut back on oil use. In , 4.6 the response of the USA to the crisis effect reveals the review of energy patterns issued by the government for the times to come. What is also interesting to note in the is the lead time in order to adapt to the new situation encountered (e.g. the natural gas contribution share started increasing 5 years after the crisis).
Another critical point concerning the weaknesses of forecasting previous to crises, not related to the use of numerical past data, may be met in the case of California. Once the regulators and the state adopted a deregulation system that was elsewhere applied successfully (Woo et al., 2003), they decided to proceed in certain modifications (i.e. partial deregulation and imposition of retail price caps) without bothering to consider the different characteristics, features and conditions of operation encountered in the California environment. Hence what might have been thought as successful elsewhere would not be a priori successful in California as well. Finally, if the modification of market structures and potential manipulations had been taken into account via the implementation of alternative scenarios assessing the risk of deregulating the Californian electricity market, certain versatile mechanisms that would instantly respond to a potential crisis may have been put forward. From the analysis provided it becomes clear that forecasting methods that solely rely on past data trends, disregard the wishes of relevant actors and major players, and do not consider the conditions forming the environment where the phenomenon develops cannot capture a broader view of the situation and thus give valid predictions.
As already addressed, the limited ability of quantitative parameters and extrapolation techniques to provide a valid forecasting, especially in the case where a crisis was to follow, is indisputable. To validate the conclusion made and further support Godet’s beliefs an example is presently given. Using the installed nuclear power data between 1967 and 1987 along with the application of extrapolation techniques (the forecast function is currently used) one may present the expected nuclear capacity time evolution for the next twenty years. A straightforward comparison of the extrapolation s with the respective real data for the period 1987 to 2007 is available in 5.1.
What of course cannot be captured by the extrapolation technique is the Chernobyl crisis, deeply influencing any further development of the nuclear installations. It was on the 26th of April 1986 that reactor number four at the Chernobyl Nuclear Power Plant, located in Ukraine exploded. By considering the magnitude of consequences that the Chernobyl accident entailed (UNDP & UNICEF, 2002), one may easily realize the cut back of nuclear capacity in the years to come. Furthermore, what is interesting to note is the different influence that the Chernobyl accident had in countries around the world. In 5.2 one may see the immediate response of the Russians, the Germans and the Ukrainians, while it took a little longer for the USA to reconsider its nuclear program. On the contrary, countries like France and Japan continued to install nuclear plants, while on the other hand Italy abandoned its nuclear program and gradually decommissioned all of its plants (NEA, 2007).
What is evaluated here, is the conditions configuring the future. Although in a global level, nuclear capacity did stagnate, this was not the case for every country. Depending on each nation’s needs, requirements and obligations, a different energy policy may be drawn. If not properly weighing these factors in the forecasting process, the outcome cannot be valid.
Based on s 5.1 and 5.2, one may also note the lead time of both the international community and the selected countries. Regarding the response of the world as a whole, a period of 3 to 4 years is to be considered for the international community to perform the actions concerned with the decision of cutting back on nuclears. As already noted, a varying response time met in different countries may be partially ascribed to the distance range from the area of the accident. However, a bundle of parameters should be evaluated in order to explain and predict each actor’s wishes, obligations and decisions.
Moreover, when investigating the long term evolution of nuclear power, one should also consider the factor of a rapidly changing environment. Since the Chernobyl accident and the stagnation of nuclear power occurred, any attempt to reestablish previous growth rates has to deal with competitors such as the galloping natural gas market, the return of the coal sector and the maturity of renewable energy technologies (Lovins, 2005). Besides, the considerations regarding waste management, decommissioning expenses and the risk of a new Chernobyl still remain strong.
Europe becoming increasingly dependent on imported amounts of energy is indisputable. According to the estimations of the recent business as usual scenarios (EC, 2007), it is expected that the energy imports’ dependency of Europe will increase from the present 50% to a total of 65% by 2030. Within this forecast, reliance on imports of natural gas is expected to increase from 57% to 84% while the respective increase for oil imports shall correspond to an additional 11%, i.e. from 82% to 93%.
In this context, European countries and Russia hold a strong interdependency bond based on the significant European energy imports of oil and natural gas supplied by Russia. Note that loss of autonomy is always a side effect of an interdependent relationship as the parties are constrained by their need for one another. Gazprom being the largest Russian company and the greatest natural gas exporter in the world (Cedigaz, 2007) constantly raises its share in the European market, with the respective volume of natural gas supplies reaching 161.5 billion cubic meters during 2006 (Gazprom, 2007), equal to approximately 26% of the total European natural gas needs. Being also Russia’s single natural gas exporter (according to the Federal Law on Natural Gas Exports adopted in July 2006), Gazprom alone utilizes the existing natural gas pipelines in order to supply Europe (see also Appendix, Existing Natural Gas Pipelines).
Meanwhile a series of recent and past events mainly suggesting disputes with Ukraine and Belarus (Bruce, 2005; Stern, 2006) have questioned the security of supply towards Europe, this revealing the potential gaming behavior of the Russians, either to satisfy political purposes or simply take advantage of the energy card in terms of increased pricing. Similar to the 1973 energy crisis and the recent oil price major increases, a scenario concerned with the raise of European natural gas supplies’ price by Gazprom is to be examined. The scenario supports that unless the desire of Gazprom for a 200% increase of natural gas prices is satisfied, supply towards Europe will be stopped.
Given the threat of a 200% price increase of natural gas heading towards European countries, an effort is presently carried out in order to investigate the measured responses of both an energy supplier and an energy user being involved in the potential crisis occurrence. Because of the particular features attributed to the subject under investigation, several cases of different energy suppliers and users may be examined. A macroscopic approach may consider two major sides, i.e. the European countries and Gazprom (Russia). However, a closer look focusing on country level and considering organizations as well is thought to be essential in order to better evaluate the situation. As already seen in the previous question concerned with the nuclear power evolution, not all countries responded in the same way to the Chernobyl crisis (NEA, 2007). Working on a country level, energy users will derive from the main Gazprom customers in both Western-Central Europe and the Commonwealth of Independent States (CIS)-Baltic countries (see also Table 6.I). On the other hand, the major energy supplier shall refer to either Gazprom or another natural gas supplier. The alternative of considering different energy sources’ suppliers will be also outlined. Furthermore, both conservative and more extreme solutions responding to the problem will be considered.
Table 6.I: Key s o