• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>IAEE International Conference;International Association for Energy Economics >DISAGGREGATING THE DIFFERENCES IN LONG-TERM ENERGY FORECASTS: THE IMPACTS OF METHOD, MODEL AND SCENARIO ASSUMPTIONS
【24h】

DISAGGREGATING THE DIFFERENCES IN LONG-TERM ENERGY FORECASTS: THE IMPACTS OF METHOD, MODEL AND SCENARIO ASSUMPTIONS

机译：消除长期能源预测中的差异：方法，模型和情景假设的影响

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

OverviewOver the years, a vast body of global energy forecasts has been accumulated by authoritative internationalorganizations, such as OECD, IEA and OPEC, as well as by respected private energy consultants, such as WoodMackenzie. These forecasts are regularly updated to reflect changing view of inputs such as policies, technologies,macroeconomic and geopolitical trends.However, the changes and forecast accuracies are rarely benchmarked and traced back to the inputs. In addition,forecasts are rarely analyzed on a long-term scale, and the reviews are largely limited to national forecasting studies,such as Craig et al (2002) for the US or Trutnevyte et al (2016) for the UK, or technology studies, such as Zheng etal (2017). In the meantime, having reliable long-term global energy forecasts is essential to underpin ambitiouspolicy targets, such as those fixed in national energy strategies under COP21.MethodWe build on a vast literature on forecast decomposition to apply three different approaches to most-cited globalenergy market forecasts: IEA, OPEC and EIA. We decompose projections for primary energy demand, as well as forcrude oil and natural gas consumption using approaches of unbalanced panel regression with fixed effects andlagged dependent; structural forecast error decomposition (Waggoner and Zha 1999); dynamic factor analysis (Stockand Watson 2002); and forecast time difference decomposition (Mason et al 2016). The decomposition targetsforecast differences due to three groups of variables. First one is external assumptions for primary energyconsumption, including socioeconomic variables (population, urbanization, GDP, car fleets and structure, housingfloors). Second one is mismatches in inter-fuel competition assumptions and structure of the energy trade, such asthe prices. We use real-time database for 2005-2016 long-term forecasts for 2030-2040 including externalassumptions and projected primary energy demand and trade by fuel.ResultsWe show that differences between the forecasts almost could not could be explained by external assumptions, suchas population, urbanization, GDP and car fleet size assumptions. Long-term price assumptions also did not prove tobe much influence in explaining primary consumption, though they strongly impacted natural gas consumptionoutputs. While variation of weighted price levels of 1% contributed around 2% to dynamics of primary energydemand at the forecast endpoint, it could drastically change the role of fuel in the resulting energy mix. The resultswere generally consistent in the approaches used. The most difference due to external assumptions is achieved withEIA and OPEC, while the implied difference due to the model structure is surprisingly largest between EIA andIEA.ConclusionsThe review of socioeconomic assumptions (both as invariant trends and storylines) was not likely to explain muchdifference in influential long-term forecasts. Given apparent low price elasticity of consumption forecast, it isplausible to assume the effect of external assumptions on the primary energy demand to be very limited in allinfluential forecasts. This is comforting given low reliability of long-term forecasts. However, where consumptionstructure is considered, forecast outputs are very elastic in the assumed long-term price levels, thus implying verylarge role of long-term price outlook as an instrument of policy parametrization. This is, again, hardly surprising,given any LCOE/LACE investment calculation is based on the price. Thus, caution is needed when approaching anylong-term energy mix forecast.
机译：概述 多年来，权威的国际机构已积累了大量的全球能源预测 组织，如经合组织，国际能源署和欧佩克，以及受尊敬的私人能源顾问，如伍德 麦肯齐。这些预测会定期更新，以反映政策，技术， 宏观经济和地缘政治趋势。 但是，更改和预测的准确性很少作为基准并可以追溯到输入。此外， 很少对长期预报进行分析，而评论很大程度上仅限于国家预报研究， 例如美国的Craig等人（2002）或英国的Trutnevyte等人（2016），或技术研究，例如Zheng等 al（2017）。同时，拥有可靠的全球长期能源预测对雄心勃勃至关重要 政策目标，例如COP21国家能源战略中确定的目标。 方法 我们以大量的关于预测分解的文献为基础，将三种不同的方法应用于被引最多的全球 能源市场预测：IEA，OPEC和EIA。我们分解了对一次能源需求以及对 使用具有固定影响的不平衡面板回归方法和 滞后依赖结构预测误差分解（Waggoner and Zha 1999）;动态因素分析（股票 和沃森（2002）;并预测时差分解（Mason et al 2016）。分解目标 预测由于三组变量而引起的差异。第一个是一次能源的外部假设 消费，包括社会经济变量（人口，城市化，GDP，车队和结构，住房 楼层）。第二个是燃料间竞争假设和能源贸易结构的不匹配，例如 价格。我们将实时数据库用于2005-2016年的2030-2040年长期预测，包括外部 假设以及预计的一次能源需求和燃料贸易。 结果 我们表明，预测之间的差异几乎无法用外部假设来解释，例如 作为人口，城市化，GDP和车队规模的假设。长期价格假设也无法证明 尽管对天然气的消费产生了重大影响，但在解释一次消费方面还是有很大的影响 输出。 1％的加权价格水平变化对一次能源的动态贡献约2％ 预测端点的需求，可能会大大改变燃料在最终能源结构中的作用。结果 在所使用的方法上通常是一致的。由于外部假设，最大的差异是通过 EIA和OPEC，而EIA和OPEC之间由于模型结构而导致的隐含差异出人意料地最大 IEA。 结论 对社会经济假设（既是不变的趋势又是故事情节）的审查不太可能解释太多 有影响力的长期预测的差异。鉴于消费预测的价格弹性明显较低，它是 合理地假设外部假设对一次能源需求的影响在所有方面都非常有限 有影响力的预测。鉴于长期预测的可靠性较低，这令人感到安慰。但是，哪里消费 考虑到结构，在假定的长期价格水平上，预测输出具有很大的弹性，因此暗示 长期价格前景作为政策参数化工具的重要作用。同样，这不足为奇。 鉴于任何LCOE / LACE投资都基于价格进行计算。因此，在接近任何物体时都需要谨慎 长期能源组合预测。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号