$5-10 trillion "now" to eliminate energy poverty

Let me explain the title further before getting into the calculations. At least $5 trillion but more likely closer to $10 trillion needs to be invested in a diverse set of energy infrastructure across many countries to bring per capita energy consumption of every global citizen to 2018 world averages. These estimates amount to about 6-12% of global GDP in 2018.

As can be seen in the following table, most of this investment will be in the electric power sector with only up to $1 trillion in the liquids infrastructure. Clearly such large-scale investment cannot happen overnight. Realistically it will take 5-10 years to build various pieces if the countries started building today.

Also, this estimate does not include the cost of replacement of aging units, the cost of new capacity to meet the demand of growing populations and economies, or the cost of building widespread, well-integrated transmission and distribution (T&D) networks.

As such, it is a very conservative estimate. But these calculations still help us define a lower boundary of the investment challenge. This is important because energy discussions today are dominated by energy transition (at least in the West) and energy poverty seems to be addressed only with initiatives such as SDG 7. The immensity of the investment estimates here suggest that SDG 7 targets are not ambitious enough (more on that below).

This investment estimate is a back-of-the-envelope calculation, perhaps slightly a bit more demanding, to have a sense of the dollar amount associated with eliminating energy poverty. As such, there are many simplifying assumptions:

• Eliminating energy poverty is simply defined as everyone reaching the world per-person average consumption levels: 3,700 kWh of electricity per year and 0.01 barrels per day (BD).

o The global mean electricity consumption in 2018 was 3,700 kWh but the global median was 2,100 kWh. I picked the mean assuming that electrification policies will continue to take hold even in low-income countries. When compared to China’s 4,600 kWh and EU’s 6,000 kWh, 3,700 kWh seems a reasonable target for eliminating energy poverty. Also, this per-capita average captures all electricity use, i.e., household, commercial, industrial and public. This is important to reflect overall economic prosperity of a country.

o The global mean liquids consumption in 2018 was 0.02 BD but the global median was 0.01 BD. Assuming electrification trends as well as increased fuel efficiency, I selected to use the median, 0.01 BD as the target. Again, this number includes all liquids use not just household consumption.

o Note that SDG 7 target of universal access to electricity does not have a consumption target. Connection to grid or having a solar panel is considered as access. But in many cases, such access provides limited and/or unreliable energy only sufficient for a few watts of electricity for a few hours a day. For example, see

§ “Real” Electricity Still Comes from the Grid

§ Realities and Challenges of Energy Access in India

o Similarly, SDG 7 target of universal access to clean cooking fuels is silent about a consumption target.

o SDG 7 does not include energy needs for transportation, commercial, or municipal service activities.

• To calculate generation investments for different technologies, I assumed CAPEX and utilization (or capacity factor, CF) numbers from the literature (see table above). Generally speaking, I picked the high end of the CF ranges and closer to the low-end of the CAPEX ranges for renewable energy technologies. The lowest CAPEX estimates are difficult to justify when building these facilities in best resource locations in low-income countries with limited transportation infrastructure, shortage of skilled labor and the need to import almost all equipment to name a few of the many obstacles. Similarly, the CAPEX for CCGT is 20-30% higher than lowest costs seen in the West.

o Note that, in reality, a portfolio of generation technologies will be built in any country. Existing generation mix reflects this reality to a certain extent. Although it may not be maintained everywhere but it is one option for the future portfolio. Single-technology options are useful in giving us a sense of relative costs of different energy policies.

• Transmission cost estimates are unsophisticated but I did not want to ignore the grid, without which sufficient amounts of electricity cannot be delivered 24x7 reliably. The weak link is that the miles-per-MW assumptions are probably too low for many locations. With this caveat in mind, miles are different across technologies to reflect some fundamental differences:

o Utility-scale solar and onshore wind, if built in best resource locations to maximize their CF, will likely be further away from load centers and the existing transmission system (to the extent one exists) than thermal plants. Offshore wind may be closer to load centers but the cost of building underwater transmission is higher.

o The miles for renewable energy technologies are also meant to capture other system integration costs such as backup generation or storage associated with their intermittency. Distributed resources also impose some grid costs. It is for this reason I have 25 miles for rooftop solar.

o Total cost is calculated based on $2,500/MW-mile. Overall, transmission costs account for a relatively small share of total costs. In reality, a lot more needs to be invested in the T&D grid to ensure reliable access to a diverse source of resources by all citizens.

• Natural gas infrastructure cost is the cost of building land-based LNG import terminals with sufficient capacity to supply all CCGT plants with sufficient gas to generate needed electricity and 20 miles of pipeline per 1000-MW of CCGT capacity, assuming that the CCGT plant can be located relatively close to the import terminal as it is already done in many locations. Increasing the pipeline mileage five-fold (roughly 100 miles) nearly doubles the total gas infrastructure cost to $1.9 trillion. I assumed $500 per metric ton of LNG import terminal capacity and $5 million per mile of pipeline. This estimate may be too high because $300 per metric ton seems to be closer to costs of recent regasification terminals but again the cost of building such facilities is likely to be higher in some frontier areas. Perhaps more relevant is that many countries have been taking advantage of the cheaper FSRU option to meet their natural gas demand rather than building an onshore regasification plant. Overall, gas infrastructure cost of $1 trillion may be a bit high.

• The liquids infrastructure cost is the cost of building domestic refining capacity at $25,000 per BD of capacity to meet all domestic demand from local production. It is not likely for all countries to build their own refining capacity or the refineries they built to use most sophisticated technology mix but this estimate may compensate for the lack of estimates on the cost of other infrastructure such as import terminals, product distribution networks (pipelines or trucking), and fuel stations. Also note that I ignored the possibility of upstream development of gas resources but some of the countries (especially those in Africa) has domestic gas resources that can be utilized rather than LNG imports.

Despite these simplifying assumptions, the total investment needs estimated in the range of $5 to $10 trillion are in line with other studies.

• IRENA estimated up to $130 trillion to be necessary between 2016 and 2050 (or about $4 trillion per year) to achieve a decarbonized energy system with every world citizen having access to energy.

• IEA Sustainable Development Scenario estimated $1.36 trillion per year between 2019 and 2030 to achieve all targets of SDG 7 (as reported in p 133 of Tracking SDG 7). This is only about $400 million more than the Stated Policies Scenario investment levels. Almost all of the additional investment is in renewable energy and energy efficiency.

• The World Bank (Beyond the Gap) estimates $0.3 to $1 trillion per year to be needed in low- and middle-income countries’ electricity sectors but not guaranteeing full access to energy nor sufficiency of energy for people with access. An additional $0.2 to $1 trillion per year is needed for the transportation sector but mostly in rail and other public transport in the cities.

IRENA and IEA are silent on what access means in terms of average consumption. But the language used in their reports is consistent with SDG 7 type of definition, that is not enough energy to eliminate poverty. The Beyond the Gap report from the World Bank, on the other hand, acknowledges that “increasing nominal access is not sufficient for the benefits of electrification to materialize.” Chapter 3 of the report provides an insightful discussion on electricity access, making distinction between tiers of access and policy drivers. Importantly, access is often delayed because many either in rural areas or city slums cannot afford to pay for electricity. But without electricity their chances of leaving poverty behind are very low. The report also acknowledges the importance of access to roads and transportation to alleviate poverty.

To break the vicious cycle of poverty and energy poverty may require governments and donor community to treat electricity and transport as public services at no cost to those who cannot afford them in the beginning. Easier said than done. Despite years of development assistance, progress has been slow for many reasons. Local governance and institutional deficiencies and changing priorities among donors are among the main reasons. But the topic is too important to ignore. Widespread poverty is a root cause of most of the world’s environmental, social and geopolitical problems.

As I continue to refine these estimates and start looking at certain countries in more detail, I welcome feedback, especially on methodology, assumptions and gaps.

A note on data sources: I used the international energy data from the EIA as the most complete and recent numbers for most of the countries in the world. As needed, I cross-checked or complemented the EIA data with data from other sources such as the World Bank, BP Statistical Review and various IEA reports.