Policy Research Working Paper 10517 Green Transmission Context, Rationale, and Planning Methodology Debabrata Chattopadhyay Durreh Tabassum Energy and Extractives Global Practice July 2023 Policy Research Working Paper 10517 Abstract Transmission is a key enabler of clean generation as the that brings together the critical elements in generation lines and substations need to be built first to encourage and transmission planning, including system security con- investments in generation. However, there has been limited straints as a mixed-integer linear programming problem. attention to readying the grid through upgrades of existing The model formulation attempts to strike a reasonable transmission lines/substations and expansion of the grid. balance between the technical rigor of a network model As a result, transmission has become a major bottleneck, and computational tractability. There are also important not only in developing countries, but also in their devel- implementation details such as making the planning period oped counterparts, including the United States, which has sufficiently long to elicit the value of transmission. The seen accumulation of 930 gigawatts of clean generation shadow prices of key constraints extracted from the model “queued up” waiting for transmission to be built. To priori- can be useful in prioritizing transmission projects, especially tize upgrading and expansion of the transmission grid, there if the duals of transmission capacity and carbon dioxide is a need to adopt a more holistic systemwide view from limits are combined. These issues are discussed around a a long-term perspective and develop a methodology that set of illustrative examples. It is expected that the model recognizes transmission as an enabler of clean generation. and associated discussion would provide a starting point to Such a methodology can be devised around a composite refine the model further and apply it to practical case studies generation-transmission co-optimization model. This paper to develop a holistic definition of green transmission and sets the context within which “green transmission” needs sustainable generation-transmission plans. to be viewed and further proposes a modeling framework This paper is a product of the Energy and Extractives Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at dchattopadhyay@worldbank.org and dtabassum@ifc.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Green Transmission: Context, Rationale, and Planning Methodology Debabrata Chattopadhyay and Durreh Tabassum 1 Keywords: Transmission investment, Green finance, Optimization model, Power system planning, Decarbonization. JEL Classification: L94 1 Deb Chattopadhyay is a Senior Energy Specialist (World Bank) and Durreh Tabassum is an Associate Investment Officer (IFC). The authors would like to thank our World Bank colleagues Claudia Vasquez Suarez (Lead Energy Specialist), Melania Lotti (Infrastructure Finance Specialist) ,Stephane Hallegatte (Senior Climate Change Advisor) and Sandhya Srinivasan (Senior Climate Change Specialist) for useful discussions on the topic. Claudia in particular has provided an extensive review of the paper that led to the inclusion of implementation challenges and Melania provided inputs on the green transmission criteria used by EU and IDB. The views and opinions expressed in this article are however the authors’ own and do not necessarily represent those of their colleagues or that of the World Bank Group. 1. INTRODUCTION 1.1 PRIORITIZATION OF TRANSMISSION THROUGH BETTER PLANNING IS NEEDED When it comes to decarbonization of the power sector, it is eminently clear that the primary focus ought to be on cleaning up the generation side, which includes scaling up renewable energy (RE) as well as phasing down of coal/gas plants. World Bank (2022) made this point emphatically during the COP27. There has been the desired focus initially on the RE scale up for at least a decade and increasingly on phasing down of coal over the second half of it – most prominently in some of the EU countries and the UK. The World Bank’s internal works on decarbonization pathways, accelerating coal transition (ACT: Huang et al, 2021), Country Climate and Development Reports (CCDR; World Bank, 2021-2023) and most recently the Scaling Up work (World Bank, 2023) have also done a comprehensive array of analytical work covering these aspects. This paper highlights the importance of the transmission investments in upgrading and expanding the grid as they obviously play a critical role in reorienting the network to accommodate clean generation. In a decarbonized state, the transmission infrastructure often needs to connect to remote RE hubs and has a significant role to manage its variability amid other challenges. Transmission and distribution (T&D) network systems will, for instance, need to cope with higher and less predictable load (as EV, heating etc. loads get added), deal counterflows in traditional “one way” distribution networks, require significant digitalization, and smartening up to deal with a much greater degree of variability of both supply and demand, low inertia and reactive power deficient system. These roles are now widely recognized in the literature and practical ways to deal with these issues are being implemented in countries where the share of renewables has gone up. However, the role of T&D and particularly transmission as the key and the first enabler that should be on an equal footing as generation is somewhat understated. The central policy research questions that this paper address are (a) What categories of transmission projects should qualify as “green”? and (b) What should be the methodology to assess their benefits? The process needs to start by planning generation and transmission together so that the complementarity of these two resources is fully captured, including the variable generation and load related issues that to a large extent require coordinated planning of the two. The fundamental point that coordinated planning of generation and transmission subsystems can lead to a cost-effective solution with or without any environmental considerations has, however, been well known to power system planners for several decades. Sophisticated research-grade methodologies that can do this has also been known for at least a decade. Linear/mixed-integer programming (LP/MIP) based models for generation or transmission planning models have been around for a long time, albeit the computational burden meant that the early models used a sequential method for the transmission plan to follow the generation plan. For instance, a nonlinear programming (NLP) based transmission planning model was proposed back in 1989 by Youssef and Hackam (1989) that can be solved sequentially taking a generation plan as an input. Bahinese et al (2001) and Akbar and Mina (2014) used MIP to undertake transmission planning. Leite de Silva et al (2010) have used meta-heuristic methods like the ant colony system to get around the difficulties of solving a very large combinatorial problem in a real-life system. Hemmati et al (2016) also use a particle swarm optimization (PSO) technique to solve the non-linear MIP model. Coordinated planning can lead to a better optimization of resources often reducing the need for generation resources with less capital-intensive transmission resources. The composite 2 generation-transmission planning problem has previously been addressed by Pozo et al (2013), Hemmati et al (2016) and Fürsch et al (2013). All of these approaches in one form or another decompose the problem into smaller components and iterate among the generation and transmission sub-components. Pozo et al (2013), for instance, use an electricity market clearing (pool) model to decide prices that feed a game-theoretic generation capacity expansion model, which in turn feed into a transmission expansion planning model. The PSO technique in Hemmati et al (2016) is also effectively a way to coordinate the MIP and NLP subproblems in generation and transmission planning problems, respectively. These models conclusively demonstrate that optimization models can be deployed to simultaneously optimize generation and transmission decisions. There are also iterative methods that interactively use an investment and dispatch optimization model in conjunction with a load flow-based network model. For example, Fürsch et al (2013) use a heuristic to coordinate between a market model and a load flow model which involves solving a large system of non-linear simultaneous equations. They showed how such a suite of models can be used to identify cost-effective transmission expansions in Europe for 2050 to meet the renewable target at a lower cost. However, the realization that such coordination matters even more in the context of deep decarbonization is of more recent origin. Although 100% renewable studies have been published for over a decade now, a review by Heard et al (2017) showed that only four of the 24 studies that they examined considered representation of transmission in any shape or form. There have been good studies in more recent years that do bring out the value of transmission. For instance, a study for the US system by Brown et al (2021) used hourly simulation for a 7-year period to show that inter-state coordination and transmission expansion reduce the system cost of electricity in a 100%-renewable US power system by 46% (from 135 $/MWh to 73 $/MWh), compared to a scenario where each state optimizes its own capacity. As Heard et al (2017) noted – a credible characterization of the transmission network expansion options is essential for establishing the feasibility of any high-penetration renewable electricity system. It would generally make a case for a higher level of transmission to lower system costs. The European study by Fürsch et al (2013), for instance, found that a proper optimization of transmission expansion options would see as much as 228,000 km of new transmission lines across Europe by 2050 compared to the existing network in 2012, which marks a 76% expansion of the grid. Brown et al (2021) also demonstrated that the optimal choice with all types of interconnection options (Alternating Current or AC as well as Direct Current or DC options) would entail 400 TeraWatt- kilometers (TW-km) of inter-state capacity to be built which is more than four times the sub-optimal less interconnected scenarios with AC lines only. The choice of transmission technologies, especially high-investment but flexible and high-capacity DC options, is an important issue. As Reed et al (2019) have argued in the US context that a conversion of the existing long AC corridors to high-capacity corridors may often be the more cost-effective solution that also avoids the critical problem of finding new right-of-way for building new transmission corridors. Their analysis shows that HVDC conversion in the US is the lower-cost option at >350 km and yields >50% capacity increases. 1.2 LACK OF TRANSMISSION CAN DELAY RENEWABLE ENTRY A singular focus on ‘scaling up’ renewables without adequate attention on transmission has not helped the very cause. Transmission projects are notoriously difficult to finance and get through the permitting processes. Consequently, it may take a very long time to build them if it does not get shelved permanently. As almost entirely fixed cost assets, financing these regulated assets typically owned by 3 state-owned enterprises in itself is a challenge. However, this is only one of several other challenges that include securing land, permits, technical challenges associated with difficult terrains and community support for lines that may span across hundreds of miles. These other challenges may in turn lead to significant cost and time overruns. As the power evacuation infrastructure needs to be in place before generation projects are commissioned, transmission projects must start as early as possible to allow for potential delays in permitting and community consultations because the regulatory approval process even for a small section of the line (e.g., a small number of towers or a substation) may jeopardize an entire project for years. SunZia Transmission Project, the largest clean energy infrastructure project in the US history to evacuate power from a 3.5 GW wind project in New Mexico, initiated the approval process back in 2006 and finally obtained it 17 years later in May of 2023 (Moore, 2023). In fact, as US DOE (2022) article titled Queued Up noted in 2022: “More than 930 gigawatts (GW) of solar, wind, hydropower, geothermal, and nuclear capacity are currently sitting in interconnection queues seeking transmission access, along with over 420 GW of energy storage… This is roughly the same amount of clean capacity needed to hit an 80% clean electricity share in 2030. It is also a large step towards the capacity needed to reach 100% clean electricity in 2035 under accelerated electrification, consistent with the nation’s decarbonization commitment.” Although the issues and magnitudes would obviously be different across countries, 2 the US example illustrates well the major conundrum in energy transition. The sheer magnitude of transmission both in terms of dollars needed and implementation challenges, merits a level of attention that is not yet forthcoming. A lack of foresight to put this critical enabler of decarbonization means a significant part of renewable projects are being delayed. 3 Ammann (2023) presents an in-depth analysis of the PJM system where 202 GW (as of September 2022) renewable projects accounting for 95% of the projects waiting in the ‘interconnection queue’. This has not only created great uncertainty for the RE investors but also contributed to significant escalation in project costs. Seel et al (2023) reports RE project costs since 2019 have gone up eight-fold. The Economist (2023) puts this conundrum quite eloquently in an article titled ‘Hug Pylons, not Trees’ as follows: “The trouble is that the scale of changes needed to adapt the world’s electricity grids is vastly underappreciated. Too little investment is taking place. Planning rules get in the way. And, in a deep and damaging irony, some of the biggest advocates of slowing climate change do not accept the logic that to do so requires building more.” It is important to put the focus back on transmission investments because most countries aspiring to achieve their NDC, JETP and Net Zero emissions targets will need to invest massively in significant upgrades, new transmission technologies and develop new green corridors to connect RE hubs, etc. Transmission to connect wind and solar is of recent origin and even in its first few years of their first hundred GW, India and China both encountered the challenges of getting it right. China, in particular, followed a massive scaling up trajectory with transmission seriously lagging behind generation, resulting insignificant RE rejections up to 34% in some provinces over 2010-2015 (Zhang et al, 2016). 2 Wind power development in La Guajira, Colombia, for instance, has run into highly sensitive social issues around indigenous people who live in the area of best wind resources (Araujo et al.,2023). 3 As Einberger and Tiplin (2022) from Rocky Mountain Institute aptly put it, “The Best Time to Plan Transmission Was 15 Years Ago. The Second-Best Time Is Now.” This is also reflected in a recently published article in the Financial Times: A. Mooney , “Gridlock: how a lack of power lines will delay the age of renewables - A backlog of wind and solar projects is waiting to connect to infrastructure built for another era, threatening net zero plans” Financial Times, June 11, 2023. Rob Gramlich (2023) also echoes these sentiments in his presentation titled ‘No energy transition without transmission’ and identifies the problems of transmission planning methodologies in a number of states in the US. 4 2. FINANCING REQUIREMENTS, CHALLENGES AND WAYS FORWARD It is useful to set the context first on the relative cost magnitudes of transmission to appreciate the investment need for transmission infrastructure. Transmission even in a large power system like the US with 1,150 GW installed capacity with vast geographical coverage, typically accounts for less than 20% of the delivered cost, compared to ~50%-60% for generation and ~30% for distribution. If we take a typical 15 c/kWh fully cost-reflective average tariff across all end users, the costs for the US system works out as 8.5 c/kWh for generation compared to only ~2 c/kWh for the transmission component (see Figure 1). Figure 1: Components of cost of electricity supply in the United States Source: US Energy Information Administration https://www.eia.gov/energyexplained/electricity/prices-and-factors-affecting-prices.php The relative costs would vary considerably across different systems, but a GW of generation generally would typically incur 4-5 times the cost compared to transmission as the capital cost (capex) even for a reasonably long 200 km transmission line would be a fraction of its generation counterpart. Transmission system operating costs (opex) are also relatively low characterized by low maintenance requirement, and asset lives that are longer than most generation assets, and offers better availability (e.g., ~98% or higher). Most importantly, as a well-meshed network with a strong backbone of high voltage network is put in place over a few decades, the incremental investment requirements over subsequent several decades is relatively low. This is how systems in the US and EU countries got into a “cruise” mode with only 1-2 c/kWh of average cost of supply incurred on transmission. However, as the system goes through a major transition with (clean) generation growing in very different parts of the system with high capital cost but near-zero operating costs, the generation cost component will drop, but that for transmission will grow disproportionate to its existing cost base. Each time a GW scale solar/wind firm comes up, there is an associated transmission infrastructure cost somewhere in the range of $150 million to $300 million. This includes all the costs from direct connection of the RE firm to the existing grid and strengthening of the core network, depending on the length of the line and level of upgrades needed. If we assume a median cost of $200 million, it translates into a levelized transmission cost of ~1 c/kWh. This will however be more than offset by the reduction in generation costs even before we consider any CO2 benefits. 5 Let us continue with the US example that provides the most well researched cost of supply figures. If we see the generation component going down from 8.5 c/kWh to say 6 c/kWh that represents a typical LCOE of a mix of solar and wind, but it is accompanied by a 1 c/kWh increase in transmission cost. Put differently, there are two separate cost impacts (a) overall cost will drop by a net 1.5 c/kWh (assuming no changes to distribution cost); (b) but transmission costs will increase from 2 c/kWh to 3 c/kWh or by 50%. Although the levelized cost of transmission in absolute terms increases only by 1 c/kWh, it has significant ramifications especially for developing countries. In a system like India with ~1700 TWh of total generation, this is equivalent to a transmission cost increase of $17 billion per annum.4 For example, Powergrid India, which is the largest transmission company in the country that owns the high voltage grid (400 kV and above) including all of the interstate transmission lines that are critical to ship power from RE-rich lower load states like Rajasthan to the mega load centers, has an existing total asset base of around $30 billion (in FY21) which gives some indication of the sheer increase in investments that will be needed. As an example, if half of the $17b pa cost needs to be incurred by Powergrid, it represents a 28% increase in its capital base in a year. In order for India to reach its 500 GW target (which is more than the existing generation capacity), Powergrid will need to grow from its current $30 billion capital base to over $100 billion by 2030. As the decarbonization process will continue over a period of two, three or four decades – this pace will need to be sustained potentially over 3-4 decades. Transmission investment numbers are likely be in the order of multiple billions per year – a volume of investment that is unprecedented. This begs the first question as to where would these dollars come from? Even for a massive entity like the Powergrid that has strong financials, sustaining a double-digit growth rate on its capital base over even a single decade is a significant challenge. The question over capital availability looms larger for the state transmission companies for the other half of the investment. Transmission, notwithstanding all the efforts to attract private investment, obstinately remains in the domain of public financing. To put this in perspective, the cumulative total private in transmission across all developing countries was reported as $11 billion in 2015 (ESMAP, 2015). Reliance on public financing is a common story, and yet with the exception of Africa and probably the odd exception like Türkiye, the recent focus of the MDBs have been more on cutting edge renewables, storage, hydrogen, etc., rather than grid strengthening projects. This trend is beginning to change, as we discuss later in this section, and may be a part of the solution to scale up transmission. Pivoting to a different example of an emerging market, it is clear that in countries like Indonesia RE at grand scale cannot happen without high-capacity AC/DC lines including inter-island cables being laid out first for islands like Sumatra to support the major load centers in the Java-Bali system. IEA’s Net Zero by 2050 roadmap (IEA, 2021) shows the extent of grid expansion needed globally to meet net zero which will require over 4 million km of transmission lines and over 4 TW of substation capacity to be built each year over this decade and the number will increase to over 6 million km and 6 TW, over the next decade (Figure 2). This will need to be accompanied by rapid digitalization of the networks that stand at less than 20% and will need to be tripled according to IEA. It is hard to put a cost estimate on the global effort. However, if we attach a notional 0.5 million per km on average,5 it amounts to $2 trillion to $3 trillion per year over the next three decades. Again, this is going to be a fraction of what is needed for the generation component but marks a phenomenal increase over the current state of play in network investments. 4 Analysis by one of the state regulators among others over the past few years has pointed to a cost impost of Rs 1.2/kWh (or 1.5 c/kWh) from renewables mostly linked to transmission (Singh, 2020). 5 This represents a mid-tier 400-500 kV new line all-in cost estimate using US DOE statistics. 6 Figure 2: Annual average global grid expansion (million km lines and TW of substation capacity) needed to meet Net Zero Source: IEA Net Zero by 2050 Roadmap, p. 180 In the near term, IEA estimates also show that the financing requirement for grid expansion and modernization will increase from $75 billion to $325 billion annually by 2030, which represents approximately 34% of the total investment by emerging market and developing economies (EDME) to achieve net zero in the power sector. Mobilization of private capital is critical to address this significant financing gap for energy transition. This calls for scaled up and possibly off-balance sheet investments by the public entity in the transmission sector. Many countries have been practicing several business models that optimize constrained public financing leading to efficient and modern grid systems. Three of the common ones are: i) privatization of the state-owned transmission company fully or partially; ii) public-private partnership (PPP) under long term concessions, a successful example of which will be the Philippines, which has seen multibillion dollar investment and notable decrease in the operating losses along with Senegal and a few countries in East Africa (Bookerd, 2022); and iii) independent power transmission (IPT) model similar to IPP in power generation where a single or bundle of transmission lines are competitively bid out to private sponsors, which has been a successful model in countries like Brazil, Chile, Colombia and India. In addition, there is a growing use of financing vehicles or instruments such as Infrastructure Investment Trust (InvIT) and Project Bonds for transmission lines around the globe and in emerging markets that are enabling inflow of equity investments from institutional investors and access to debt from capital markets. Investors (especially institutional investors) are attracted to InvITs as they offer predictable dividends from stable operating transmission assets. InvITs also provide private developers or state-owned transmission companies the much-needed platform to monetize their operating assets in order to free up debt capacity to undertake more development projects (e.g. PGInvIT in India).6 Similarly on the debt side, Project Bonds allow nationally strategic or large-scale transmission projects to access debt from a large pool of investors. Historical issuance volumes for Transmission Line Project Bonds have grown from approx. $1.0 billion pa in 2010 to $1.6 billion pa in 2018, with high growth in 6 Power Grid Corporation of India Limited (POWERGRID), under Ministry of Power, Govt. of India is the Sponsor of PGInvIT: https://www.pginvit.in/. 7 2016 and 2017 at volumes north of $3 billion pa as shown in Figure 3, which represents 13% of all Power- related Project Bond issuances by volume during 2010-2018. While the majority of the transmission line project bonds have been issued in developed countries, emerging markets such as Brazil, Chile, Peru and India have seen over $1 billion of investment through such bond issuance (Credit Agricole CIB, 2018). Figure 3: Total transmission project bond issuance volume (2010-2018 USD equivalent) Source: Credit Agricole CIB, 2018 2.1 CONCESSIONAL FINANCING AND THE GREENNESS OF TRANSMISSION One of the challenges of financing transmission in general, and concessional financing through MDBs in particular, centers around the “greenness” of transmission. There are seemingly many good transmission projects in any country that are needed to provide more access to remote villages, enhance reliability or improve economics, regardless of whether they receive concessional finance, or not. For examples, projects with innovative first-of-its kind HVDC applications to manage renewables in countries like Vietnam, Indonesia, Türkiye etc. should count as critical enablers and get sufficient consideration when channeling concessional funding. The “next highest voltage” class - be it HVDC or HVAC – innovative or not – is also lacking in several countries including Türkiye that has stubbornly stuck to 380 kV as the highest voltage, or Indonesia that either has a skinny 150 kV system or a limited number of 500 kV lines with almost nothing in between. Helping these countries to find a path to a robust RE-ready transmission network should make useful Technical Assistance (TA) and pilot projects. Digitalization and smartening up of transmission systems to improve resilience, enhance efficiency of transmission are missing in most of our development partner nations that should also be a serious contender of concessional financing. Smart grid options in transmission are still relatively scarce in the developing countries although roadmaps, pilots and a small subset of these technologies have been implemented in some of the systems. Technologies like HVDC, fast reactive power devices (like synchronous condensers/compensators, static reactive power (VAr) compensators), wide area monitoring systems (WAMS), substation automation, dynamic line rating (DLR), etc. to name a few, can equip the system to deal with unpredictable variability in supply and demand, and use the resources in a more efficient way. As an example, DLR which allows the system operator to vary transmission line capacity rating in real-time depending on weather conditions, has been shown to allow 30%-100% higher utilization of transmission lines. As Elia, the Belgian system operator, has demonstrated, DLRs can be particularly useful to evacuate power from wind farms because higher wind speed allows 8 transmission thermal capacity to be raised (because the heat generated by the lines can be dissipated more easily) and reduce the need to curtail wind power significantly (Schell et al, 2013). Synchronous condensers (SYNCON) provide another good example that can be useful to provide valuable reactive power (and inertia) that would normally be provided by conventional (mostly thermal) generation in a system, to enable a greater amount of (asynchronous) solar/wind generation to be integrated in the system. As Filatoff (2021) describes, two SYNCONs that have been installed in South Australia allowed nearly doubling the wind generation that could be accommodated in the system. An innovative approach that is now gaining popularity, including the World Bank’s Komati project in South Africa (Kwakwa, 2022), is to repurpose old coal-fired generators into SYNCONs at low cost to essentially retain the good part of the (ancillary) services that the coal plant used to provide. Last but not the least, conventional transmission projects for grid strengthening can also contribute to greening of power systems that need to be recognized. More often than not, a typical economic analysis of transmission, including those conducted by the World Bank, looks at the first 10-15 years of an asset that typically lasts for 50-60 years, and yet there is significant concern about its net positive CO2 emissions over these early years. This is a wrong way of looking at it as energy transition will not happen over a decade or even two decades, and the lines in the early part of its life will carry coal/gas- based power. To assess if grid investments ‘should’ or ‘to what extent’ be attributable to Climate Finance (concessional type as such GCF, IDA PSW etc.), several criteria have been put forward in practice. These include the EU Taxonomy developed by the European Commission and the Common Principles approach developed by MDBs and DFIs. The EU considers transmission and network to be green only if two-thirds of the newly connected generation capacity has CO2 emissions intensity below 100g CO2e/kwh or if the average grid emissions factor is below 100g CO2e/kwh over a rolling five-year average period (Pye, 2021). This is a somewhat restrictive, narrow and myopic view and if a more forward-looking view on transmission and critical scale-efficient transmission projects cannot be inculcated, energy transition will almost inevitably get stuck mid-way. Common Principles on the other hand uses a non-binary forward looking approach where it gives partial climate credit to grid investment based on the share of the very low carbon electricity in the grid over a time horizon such as 10 years (Pye, 2021) unless the grid lines are solely dedicated for evacuating very low carbon electricity generation in which case the total investment is fully attributable to climate finance. Nonetheless based on a recent analysis presented in CoP26, it is estimated that less than 40% of the grid investments needed in EMDEs by 2030 would be climate finance attributable under the current eligibility criteria in use. In summary, there are very useful transmission upgrades, new transmission line technologies, smart grid technologies that can support greening the transmission and should qualify as green investments that deserve concessional financing. Planners in ministries, regulatory bodies, transmission utilities, banks and MDBs – all need to consider adopting a more holistic systemwide view from a long-term perspective and develop a methodology. It is entirely possible that implementation of such a methodology will be fraught with difficulties to find model, data and capacity to undertake such analysis. Nevertheless, as the rest of this paper argues, a co-optimized generation and transmission planning framework can be put in place that captures the essential intrinsic merits of a wide variety of transmission projects including critical upgrades to the existing network. 9 2.2 ILLUSTRATIVE EXAMPLES Let us consider an illustrative example to understand how the transition may take longer but why transmission upgrades in the short-term is important to realize it. Figure 4 shows a typical system expansion that might see initially require an upgrade to the existing network (e.g., a low-capacity 220 kV line) to ready the grid for a higher penetration of RE. Figure 4: Illustrative example of green transmission upgrade opportunities Short to medium term (2023-2030) Long term (2030+) This is how the system expansion may occur over the years: • Consider an existing grid (represented by the dotted circle) on the left panel with a coal generator that is connected to the load center by a 220 kV line. Due to past load growth, the line is already at the limit requiring part of the (peak) load to be met through a local LNG-fired expensive generator near the load center. There are remote RE resources that have started developing north of the coal plant requiring a new line to be developed and a pooling station (green dot) that collects solar/wind into the RE substation. These new assets (pooling substation and new line) will clearly be ‘green’ according to the incumbent EU/IDB criterion. But the existing 220 kV line will also need to be upgraded (to say 400 kV). • The question arises on the upgrade (or it could well be a new 400 kV line) of the existing grid. It is obviously needed for the RE generation to meet load. And yet as soon as the 400 kV line is in place, the immediate impact would be an increase in coal generation (and commensurate drop in LNG) together with some RE that will come online. Net emissions will increase, and the specific contribution of the line may also be a net positive until say 2030, although the line may in fact pass the economic test with a decent benefit to cost ratio. • In the long-term (post 2030) though, the RE hub will grow and the coal plant may be retired and repurposed to become a flexibility center (Kwakwa, 2022; Huang et al, 2021) with the generator 10 converted to a synchronous condenser. The site may also install battery storage to continue to use some of the existing assets. The storage may not only provide useful balancing service but also defer further upgrades to the 400 kV line by managing the RE load near the RE hub. Emissions in the long term will decrease to meet the net zero emission goal – this is an illustration of how the transition is supposed to work. • The upgrade to the existing line to 400 kV unfortunately does not meet the definition of green transmission but it should be as it is the critical near-term enabler to ensure RE scales and coal phases down. It is in fact also critical to ensure private investment in RE happens at scale to develop the RE hub. Therefore, such transmission investments should lead the transition because it encourages RE development, reduces system cost and improves reliability. Every time we draw one of those downward sloping emissions curves for Net Zero target, there is a case for such transmission projects that need to lead that curve by a decade. It is important to do these projects on a timely basis especially in crowded countries like India and Indonesia as the transmission owner will get at best one chance to make a good use of a corridor. They should make those chances count and build high-capacity lines. It also leads us to the corollary that transmission should get low- cost capital and even a priority over RE. How this is achieved in reality will require some smart thinking though because the regulatory processes in most places including developed countries currently do not allow a preferential treatment of scale-efficient transmission to be built ahead of RE generation, and the definition of ‘green transmission’ to date remains hopelessly myopic. Justification of transmission to support energy transition in coal/gas heavy systems is particularly difficult in most of the developing countries. The vertically integrated utilities are often in a financial mess and the idea of investing billions in those institutions may not be attractive. Concessional financing in the current scenario (regulations and financial state) will require very specific arguments. One can probably find such arguments for specific cases like hydro in Nepal and the Republic of Congo or RE in Sumatra, to displace coal elsewhere. What is needed is a new regulatory framework that explicitly encourages transmission at good scale to connect RE hubs. There was such a proposal for Scale Efficient Network Extension (SENE) in Australia (AEMC, 2011; Chattopadhyay, 2011) but it seemed too bold and the Australian Energy Market Commission (AEMC) in the end ruled against it to stick to the incumbent project-specific analysis. SENE was a departure from a case-by-case project economic analysis to favor scale efficiency and supporting the current pace of energy transition needs such an approach. It is a pity that Australia did not embrace the idea as it basically led to some of the largest solar, wind and geothermal projects to disappear. If we are serious about energy transition requiring massive scale up of RE, we really cannot achieve it building one line at a time only when each RE project is committed or partly in place. There is a risk of stranded transmission asset if transmission is built at scale and RE does not eventuate. Nevertheless, if scaling up RE at the desired pace is warranted, countries have to take on that risk and do large HVDC lines, undersea cables, ultra HV AC lines before the RE projects are committed. “Build them and they may not come” is a risk but if sometime the only corridor that is available is used up, this action forecloses the option to build further RE, or substantially delay RE – it is a risk too and probably a bigger risk that may jeopardize energy transition as the discussion in the US context already alludes to. 11 3. MODELING METHODOLOGY The proposed modeling methodology relies on an LP/MIP formulation that captures the essential elements needed to address the issues discussed. The model is cast as a single LP/MIP optimization to represent the multi-year security constrained generation-transmission planning problem. It tries to strike a balance between the granularity of the network technical details with the broader objective of solving a multi-year carbon-constrained intertemporal optimization problem. The former objective is important to recognize some of the unique properties of a transmission network – most notably the interdependence of the flows in an AC network. The latter is, however, critical to prioritize the transmission projects that best serve the objective of meeting a progressively tightening CO2 limit. The model, therefore, deviates from a nonlinear load-flow centric methodology but retains a linear “DC” approximation of the power flow constraints. The intertemporal optimization captures these network constraints together with other transmission, generation, system security and carbon constraints. The methodology resembles the modern power system planning tools like PLEXOS Long Term module 7 and the World Bank Electricity Planning Model (EPM; Chattopadhyay et al 2018) among others. The intrinsic value of transmission to decarbonize a power system needs to be brought out through a model that: (a) Has the right set of features including the “look ahead” to make the trade-off over time as well as across generation and transmission resources. The model needs to recognize the interdependence of flows in a network that in turn means interaction among transmission projects, e.g., flow on a green corridor will change those in other existing fossil fuel heavy transfers and vice versa. The role of spinning reserve to maintain system security in a RE heavy system is also important to consider because it affects dispatch and hence flows on lines around the network typically increasing the need for transmission capacity; (b) Is implemented correctly including an adequately long planning horizon to elicit the value of transmission projects that are necessary to retire existing thermal generation, bring in cleaner forms of generation at a rapid scale, represent the requisite firm capacity margin as well as operating reserve, and stability/inertia related constraints as best as possible. It would also be important to present the data to elicit the scale-efficiency of transmission options by giving multiple alternatives at different voltage class as well as AC vs DC options, and realistic gestation lag of transmission projects; and (c) Interpret the results of the model carefully including appropriate usage of the shadow prices to prioritize projects. Section 3.3 provides some insights into these issues. This section provides a mathematical exposition of the model variables and equations. 8 7 Features of the PLEXOS model are available online: https://www.energyexemplar.com/plexos. 8 It should be noted that a full-blown planning model has many features that would run into at least 25 pages and many of those are peripheral to the topic of green transmission. The core features of the model are discussed in sections 3.1-3.3. Chattopadhyay et al (2018) and the EPM model on GitHub present the full details: https://github.com/worldbank/EPM. 12 3.1 NOTATIONS Symbol Description Units Sets G Generators Y Years i,j Nodes/zones/substations D Representative days T Hours , Generator to node mapping , Connectivity matrix of nodes Generators critical for stability/inertia , Lines critical for stability/inertia Input parameters , Annualized generation capex plus fixed $/MW/year operations and maintenance costs ,, Annualized transmission capex $/MW/year , Variable fuel and operations maintenance cost $/MWh ,,, Hourly demand for each representative day MW , Transmission capacity of line (i,j) MW , Equivalent line impedance Per unit Ohm , Number of hours in day type d for block t Hours , Annual max availability of generator g in y Hours , Minimum run requirement for g in y Hours , Ramp up and down rate for generator g MW/hours ,,, , ,,, Spinning reserve up and down requirement MW Local reserve margin Fraction , Annual peak demand MW ,, Stability limit MW and tons Annual CO2 limit ,, and ,,, Weight on generation and transmission variable Annual limit on power sector CO2 emissions Tons , Loss factor Fraction Decision variables , and , Generation capacity and cumulative generation MW capacity at g ,, and ,′ Transmission line investment decision and 0 or 1 cumulative value ,,, Generation dispatch for day type d and block t MW ,,,, Directed flow from node i to node j MW ′,,,, Undirected flow from node i to node j MW (+ or -) ,, Unserved energy at node i MW Power angle at node i Radian ,,, Spinning reserve – up/raise (to increase MW frequency) ,,, Spinning reserve – down/lower (to increase MW frequency) ,,, and ,,, Unserved up/down spinning reserve MW 13 3.2 GENERATION AND TRANSMISSION CO-OPTIMIZATION PLANNING MODEL The basic construct of a composite generation and transmission planning optimization is as follows: Minimize the discounted value of total cost where the undiscounted cost function is defined as the total cost of generation and transmission investments and variable cost of generation: = � , . , + � ,, . , . ,, + � , . ,,, , ,,′ ,, , Subject to three sets of constraints associated with transmission, generation and system level, namely: Transmission related constraints: 1. Demand at each node of the network must be met by either dispatching generation located at the node or through import barring the possibility of unserved energy that cannot be met: � ,,, + � ,,,, − � ,,,, . , + ,, = ,,, ∈, ∈, ∈, 2. Transmission capacity over and above existing capacity (ICAP’) will need to be co-optimized together with generation decisions. Transmission decisions (IT) for a corridor will typically be a binary variable although a continuous approximation allowing IT to vary between 0 and 1 for aggregated zonal representation for multiple lines may also be appropriate. Note that CIT is a cumulative variable that in a linear model may allow staging of transmission projects: ,,,, ≤ ,′ . , + ′, ∀ (, ) ∈ , ,, = −1,, + ,, ∀ (, ) ∈ , � ,, ≤ 1 , ∀ (, ) ∈ , 3. Transmission flow from node I to node j is defined as the difference between the power angles at the two nodes. This is the “DC approximation” to power flow constraints that effectively assumes there are no reactive power limitations in the network allowing voltages at the nodes to be at their nominal values (of 1 per unit). It should be noted that the free or ‘undirected’ flows will be positive in one direction and negative in the other direction. Since the model accounts for losses on flows (in the demand balance), the undirected flows are split into two positive directed flows in each direction: ( − ) ′,,,, = ∀ (, ) ∈ , , ′,,,, = ,,,, − ,,,, ∀ (, ) ∈ , 14 Generation related constraints: 4. Generation dispatch variables are constrained by capacity (existing capacity GCAP’ plus new capacity IG that are built over the planning period) as well as maximum availability hours (net of any planned and forced outages). It will also be constrained to observe a minimum number of hours that it may need to run (MINRUN) either to meet contractual obligations (which strictly speaking should not be part of an economic analysis) or as an approximation of inflexibility that thermal plants typically exhibit with a minimum up time constraint once the boilers (for coal/nuclear) or gas turbines have been fired up. Finally, thermal generators will also need to abide by the ramp up and down rates from one hour to the next: ,,, ≤ , + ′=1, , = −1, + ,, � ,,, . , ≤ , . , , � ,,, . , ≥ , . , , ,,, − ,,,−1 ≤ ,,,−1 − ,,, ≤ System reliability, security and emissions related constraints: The system level constraints differ in terms of their scope – 5. The generation reliability constraint stipulates that the total installed capacity in any year must exceed the annual peak demand in a node (i.e., the highest value of ,,, ) by certain margin (LRM or local reserve margin), or capacity reserve. This constraint for a small country/system may be applied at the system level too. The LRM parameter will depend on a number of factor including uncertainties around random outage of generation as well as demand: � , ≥ (1 + ) . , ∈, 6. In addition to capacity reserve that renders the system reliable from having enough firm capacity, there is also a need for operating or ‘spinning’ reserve at a number of generators that are typically capable of delivering such reserve rapidly in the event that a contingency event has occurred. Spinning reserve may be in the “up” direction (SRUP), i.e., generators will need to increase their production because generation or transmission line has been lost (or demand has spiked), or the reverse situation may occur if a load has been lost and spinning reserve down (SRDN) may be in the form of some generators rapidly lowering its generation level. SRUP 15 requires some generation capacity must be freed up so that the generator can take on additional load (up to its overload capacity reflected as a share of capacity or ): ,,, + ,,, ≤ , . (1 + ) Spinning reserve down, on the other hand, requires there is enough generation allocated on the generator for it to lower its output: 9 ,,, ≥ ,,, Spinning up and down requirements may need to be met at a nodal or system level depending on the characteristic of the system and associated regulatory requirements. The spinning reserve up and down requirements (SRUD and SRDND) � ,,, + ,,, ≥ ,,, ∈, � ,,, + ,,, ≥ ,,, ∈, 7. In addition to capacity and operating reserve requirements, there will typically be a range of dynamic stability, minimum inertia and voltage limit and stability constraints that are often approximated as linear constraints in the dispatch optimization (see for instance, Transpower, 2012). These constraints will generally combine the dispatch (G) and flow (F) variables with different weights on each of them to reflect how stability and inertia can be preserved in the system. For instance, generators near load centers may often be needed to stay on above certain limit to ensure there is supply of reactive power locally to keep voltage above a minimum level, or there may be a group of such generators and import flows over AC lines that are needed to maintain sufficient inertia: � ,, . ,,, + � ,,, . ,,,, ≥ ,, ∈ ,∈, 8. Finally, there may be a maximum annual limit that represent a pathway to meet the carbon policy objective: � ,,, . 2 . , ≤ ,, 9 The constraint in actual operation of the power system/market would typically be represented with integer variables as there would be a minimum generation level below which the unit cannot be operated. There may also be specific zones between the minimum and maximum generation levels when a unit is capable of offering frequency response. Some of these complications cannot be represented in full details in a long term planning model due to the size of the model that needs to cover multiple years, hundreds of generators across many days and hours. 16 3.3 PRIORITIZING TRANSMISSION LINES USING SHADOW PRICES Shadow prices of the transmission capacity and CO2 limit can provide good insights to prioritize transmission projects. This is obtained as follows: • A=δZ/ δCO2, is the dual of the CO2 limit (i.e., shadow price of the CO2 cap constrain) and Bi,j= δZ/ δ(ICAPi,j), where o i and j are the nodes o Z is the total system cost comprising generation and new transmission build costs (objective function) o CO2 is the CO2 cap o ICAPi,j is the transmission limit in the constraint: ,,,, ≤ ,′ . , + ′, • Bi,j for new links represents the marginal investment needed to expand the capacity by 1 MW. • A/Bi,j = δ(ICAPi,j )/δ CO2 yields the marginal transmission investment needed per ton of CO2 reduction which can be normalized by dividing with the relevant number of hours. In our example, we have assumed one average load condition and hence it can be divided by 8760 hours to give us the marginal MWh capacity needed for an equivalent ton of CO2 reduction. • Note that the ratio (A/Bi,j) simply relates the shadow price of transmission flow limit with that of CO2 cap which makes the comparison across new transmission candidates easier. One would be interested to look for paths that have the lowest ratio. 3.4 A NUMERICAL EXAMPLE AND DISCUSSION ON METHODOLOGY Let us consider a simple example using the proposed planning model, to make a call on the greenness of transmission (or lack of it as the case may be). It builds on the example that was discussed in Figure 3. Figure 5 shows a simple power system that has four zones (A,B,C,R), including, 1. the existing fossil fuel dominated grid comprising A-C that have coal generators in A-B and an expensive gas generator at C, and all the loads including the largest load at C; and 2. a large solar hub R that will need new lines to be constructed to A-C to meet the load. 17 Figure 5 Numerical example to illustrate green transmission Notes: Exp. cost refers to transmission expansion cost in $ million per MW. There is no limit imposed on line expansion capacity. Losses on all lines are assumed to be 1%. It should be noted that the total load of 11 GW exceeds the existing thermal generation capacity although the transmission capacity of the existing grid (AB, AC and BC) is more than enough to meet the largest load at C. Expansion of transmission capacity is, however, inevitable to meet load growth as RE is assumed to be the only generation expansion option allowed in our example. Node R can either be connected to B via a short line costing $0.1m/MW or A for $0.2m and a direct connection to the largest load node C comes at a substantially higher cost of $0.5m/MW. The example tries to capture a planning scenario wherein a restriction on new build coal/gas is imposed and further there is a 10 mtpa CO2 emissions cap (as compared to a current policy scenario or CPS emissions level of 77 mtpa). How do we decide on solar capacity addition, transmission connection and what do these choices entail for greenness of various transmission corridors? As the model in section 3.2 discussed, we need a planning model ideally capable of co-optimizing generation and transmission investment and operations decisions, i.e., it needs to simultaneously pick the solar capacity as well as the best points of connection, flows, and dispatch of generators to meet the loads (Lew et al, 2022). One of the methodological challenges that we will conveniently ignore with our toy example is that such a problem can be computationally very demanding. While such planning models exist, the computational burden for a real-life system with hundreds if not thousands of generators and lines for 20-30 years can render the model to be extremely large with tens of millions of 18 variables and constraints. In this case of course we have only 4 generators, maximum of 6 lines, and just one year with a single time period corresponding to an average load condition for all 8,760 hours. Table 1 presents the salient outcomes of the two scenarios: (a) Current Policy Scenario (CPS) that does not allow any further coal/gas build but does not restrict CO2 emissions; and (b) Accelerated Decarbonization scenario (ADS) that imposes a 10 mtpa limit on CO2. Table 1 Comparison of scenarios: High RE development Parameter Current Policy Scenario Accelerated Decarbonization Scenario Annual cost $m 4432 (Generation 3913, Tx 519) 6850 (Generation 5656, Tx 1149) Generation decision 3.1 GW of solar 10 GW of solar Dispatch GW A = 4 GW, B = 4 GW, R = 3.1 GW A = 1 GW, C = 0.1 GW, R = 10 GW New transmission GW BC 2.1 GW, RB 3.1 GW BC 1.9 GW, RB 10 GW Flows GW AC 2 GW, BC 6.1 GW, RB A B C 3.1 GW A 2.0 B 2.9 5.9 R 10 LMP* ($/MWh) A 72.9, B 72.1, C 84.4, R 60.0 A 146.9, B 145.5, C 158.5, R 132.6 CO2 emissions (mt) 77.1 10 Shadow price of CO2 N/A 116.9 * Locational marginal price that includes generation and transmission investment costs. There are four key aspects that stand out even in such a simple example: 1. Total annual cost increases sharply as there is a very steep emissions reduction target – generation costs increase by 40% but the transmission component more than doubles (albeit in absolute terms generation cost increases more); 2. In both scenarios, the renewable hub (R) connects into node B as this is the cheapest line to build and this in turn means the flow on path BC increases requiring the existing interconnector to be upgraded. As an aside, it is worth noting that if the new transmission solution to RC is forced in this example, it will marginally reduce generation cost from $5,656 million to $5,543 million but transmission investments would increase from $1,149 million to $4,988 million. Therefore, an upgrade to the existing grid in this instance leads to a very significant savings and an upgrade to the incumbent BC path which is used to evacuate coal generation now also becomes an integral part of the decarbonization strategy. It is also worth noting that the (public) investment in BC of $190 million (1.9 GW at $0.1 million/MW) allows ~$10 billion ($1 million/MW of RE) in (private) generation investment. If the BC upgrade option is eliminated, the RC capacity will need to expand by ~2 GW at an additional annualized cost of $570 million will discourage at least part of the generation investment; 3. The locational marginal prices (LMP) in ADS more than double in part because the expensive gas generator in node C sets the price, but also because of the increase in transmission investments that in a co-optimized generation-transmission planning model reflects the marginal investment needed in the network; and 4. Finally, we present a potentially useful metric that could be elicited by dividing the duals of the transfer flow limit by that of the CO2 emission cap following the preceding discussion in the last 19 section of the methodology (section 3.3). As the table below shows, RB and BC paths have the lowest MWh per ton of CO2 reduction ratio, that indeed formed part of the optimal expansion: A B C A 0.085* B 0.098 R 0.195 0.098 0.488 * This path upgrade has not been allowed and as such even if such an upgrade would have lowered the cost of meeting the emissions target, it is not relevant in this particular case. A more tedious but comprehensive alternative would be to evaluate each transmission expansion alternative as we have in fact done for the RC alternative and discussed before. The significance of the existing grid upgrade (BC) also needs to be appreciated in light of the uncertainties around RE development. Let us consider an alternative version of CPS in Table-2 wherein only 2.5 GW of RE is developed and a variation on it when the BC link is dropped. Table 2 Comparison of Current Policy Scenarios: Low RE (2.5 GW) development Parameter Current Policy Scenario CPS without BC link Annual cost $m 4514 (Generation 4116, Tx 397) 4883 (Generation 4247, Tx 636) Generation decision 2.5 GW of solar 2.1 GW of solar Dispatch GW A = 4 GW, B = 4 GW, C=0.6 GW A = 4 GW, B=4 GW C = 1 GW, R = 2.1 GW R = 2.5 GW New transmission GW RB 2.5 GW BC 1.5 GW RC 1.1 GW, RB 1.0 GW Flows GW B C B C A 2000.000 A 2000.000 B 4000.000 B 5475.000 R 1010.101 1070.707 R 2500.000 LMP ($/MWh) A 88.4, B 87.5, C 100.0, R 75.3 A 71, B 72.1, C 118.3, R 60 CO2 emissions (mt) 79.7 81.5 Shadow price of CO2 N/A N/A As Table 2 shows, removing the BC link leads to sub-optimal outcomes including: 1. It actually impedes RE development with 2.1 GW solar developed as opposed to the 2.5 GW; 2. Although both generation and transmission investments are higher to build RC link; 3. There is more local gas generation needed which not only increases cost but also leads to higher CO2 emissions. If the two sets of observations from Table 1 and Table 2 are combined, it is eminently clear that the upgrade to the existing grid is part of what would constitute a ‘robust’ solution that is needed at some level regardless of the RE development or the CO2 reduction scenario. It imparts a sense of assurance to RE developers who do not need to wait on construction of long/expensive and difficult transmission 20 links (like RC), reduces system cost and indeed more capacity on a critical path renders the system more secure. Although we do not discuss the system security benefits in these examples, it is in itself a strong reason to upgrade the transmission network to cover for contingencies such as sudden loss of thermal or RE generation.10 In fact, a fair share of ‘reliability entry’ transmission has been justified solely on the ground of meeting the (n-1) security criterion. If we want to make energy transition a reality, we must find a similar criterion to give a shape to the sustainability criterion. 4. CONCLUDING REMARKS In any discussion on decarbonization policies, there is an instant and unanimous recognition that transmission is an integral part of the solution for energy transition. However, in reality there has been a far greater focus on scaling up renewable generation and more recently phasing down of coal. Transmission, which is the critical link between these two to reorient the network to facilitate delivery of renewable power, seems to have fallen through the gap, or perhaps it has been taken for granted that the requisite infrastructure will be financed and put in place before RE investments take place. Given the serious implementation challenges associated with transmission that have been encountered even in the advanced economies, it is paramount that developing nations with their weak networks avoid the situation in countries like China that experienced severe RE rejection at great cost, or the US that has spent 17 years on its largest transmission project and 930 GW of generation overall waiting for transmission approval. Transmission has historically been justified more on the grounds of reliability rather than economics. It is time to recognize that it is a critical enabler of clean generation to combat climate change and give the relevant grid upgrade, expansion and modernization projects the push that they deserve. In order to decide what is a relevant project, it is possible to use existing models, data and tools that (a) co- optimize generation and transmission; and (b) take a sufficiently long “look ahead” to capture the long- term benefits of the network. As we have illustrated using a set of simple examples, the relevance of grid strengthening projects that are not directly connecting a renewable project can then become quite evident. Transmission planning needs to be integrated into the generation planning models. The academic literature has identified this need for at least a decade and there are also insightful applications that have come about in the last few years that highlight the need for such a model to underline the role of transmission in decarbonization of power systems. It is, however, important that some of the core features that are critical to elicit the value of transmission are given prominence without destroying the transparency of the model or rendering it computationally intractable. In order to implement such a methodology, planners will inevitably come across data and computational challenges. It may also mean making assumptions on network connection, disaggregated demand, etc. that are somewhat outside the incumbent planning norms in most places. Nevertheless, the significant role transmission plays can only be appreciated through a long-term 10 There is a limited but growing literature on the criticality of transmission and specifically the role of transmission planning. The paper by Lew et al (2021) has an excellent discussion on US and European case studies that show how transmission can help in lowering system cost significantly (including a vast reduction in need for storage) albeit transmission itself remains a small part of overall system cost. The paper also presents good arguments on the role HVDC transmission can play. 21 analysis. The thoughts and analysis presented in this work will require significant additional efforts to fine tune the methodology and implement it for a real-life system to test the veracity of the approach and what processes need to be in place to make reasonable assumptions. As the transmission grids in the majority of the developing nations are already quite weak with inadequate number of high voltage lines, modern transmission technologies and very few smart grid components, investments in voltage upgrades, expansion and modernization hold the key to decarbonization. As transmission projects are notoriously difficult to implement with long delays over permitting, environmental, regulatory and right-of-way clearances, it is time critical for MDBs to prioritize transmission projects in developing countries so that these can eventually pave the way for renewable generation in remote areas so that these countries can avoid what has already happened in the US and China. One of the stumbling blocks for this to happen is the myopic definition of green transmission that effectively qualifies only direct connection assets to clean generation projects to be funded. Such a restrictive definition rules out the vast majority of the grid strengthening and smart grid components that are quite central to the ability of the grid to carry clean power from the connection points to the load centers. It is envisaged that the methodological framework presented in this paper will provide a starting point for refining the model through practical applications and developing a more holistic definition for green transmission. 22 REFERENCES AEMC. 2011. Scale Efficient Network Extension: Final Determination, Sydney, Australia. https://www.aemc.gov.au/sites/default/files/content/6073134e-f3de-4af8-8ac9-855f6866897a/Final- Rule-Determination.pdf A.M. Leite da Silva, L.S. Rezende, L.A. da Fonseca Manso, L.C. de Resende. 2010. Reliability worth applied to transmission expansion planning based on ant colony system, International Journal of Electrical Power Energy Systems, vol. 32, pp. 1077-1084. P. Boonkerd, 2022. Attracting private finance to transmission in the Asia-Pacific Region, United Nations Economic and Social Commission for Asia and the Pacific. Thailand. Available online: https://policycommons.net/artifacts/2473823/attracting-private-finance-to-transmission-in-the-asia- pacific-region/3495880/ B.P. Heard, B.W. Brook, T.M.L. Wigley, C.J.A. Bradshaw. 2017. “Burden of proof: A comprehensive review of the feasibility of 100% renewable-electricity systems” Renewable and Sustainable Energy Reviews, vol.76, pp.1122-1133. Credit Agricole CIB, 2018. Project Bonds: Power Transmission Lines. Available online: https://www.ca- cib.com/sites/default/files/2018-11/Project%20Bond%20Focus%202018%20- %20Transmission%20Lines%20_%20FINAL.PDF D. Ammann, Waiting Game: How the interconnection queue threatens renewable development in PJM, May 2023, Report#R23-04-B, NRDC. D. Chattopadhyay. 2011. Scale Efficient Network Development to Support Renewable Generation Development, IEEE Transactions on Sustainable Energy, vol.2(3), pp.329-339. D. Chattopadhyay, F. de Sisternes and S. Oguah. 2018. World Bank Electricity Planning Model, World Bank, Washington DC. Available online: https://www.researchgate.net/publication/325534590_World_Bank_Electricity_Planning_Model_EPM_ Mathematical_Formulation_World_Bank_Electricity_Planning_Model The model source code in GAMS and documentation is available: https://github.com/worldbank/EPM D. Lew et al. 2021. Transmission Planning for 100% Clean Electricity, IEEE Power and Energy Magazine. D. Moore. 2023. The Bitter 17-Year Saga to Build a Power Line Critical to US Climate Action, Bloomberg, May 2023. Available online: https://www.bloomberg.com/news/features/2023-03-15/why-a-us- transmission-line-planned-17-years-ago-still-isn-t-built#xj4y7vzkg D. Pozo, E.E. Sauma, J. Contreras. 2013. A three-level static MILP model for generation and transmission expansion planning, Power Systems, IEEE Transactions on Power Systems, vol.28 , pp.202- 210. Economist, Hug pylons, not trees: The case for an environmentalism that builds, April 2023. 23 ESMAP. 2015. Private Sector Participation in Electricity Transmission and Distribution: Experiences from Brazil, Peru, the Philippines, and Turkey, Washington DC. Available online: https://www.esmap.org/node/55981 European Union, Sustainable Finance Taxonomy: Activity 4.9 Criterion #1.2022. Available online: https://ec.europa.eu/sustainable-finance-taxonomy/ H. Youssef, R. Hackam. 1989. New transmission planning model, IEEE Transactions on Power Systems. Vol. 4, pp.4 9-18. IDB, Green Transmission. 2022. Available online: https://www.idbinvest.org/en/green-transmission-line International Energy Agency, (2021a). Net Zero by 2050: A roadmap for the global energy sector, October 2021. Interactive version available online: iea.li/nzeroadmap International Energy Agency, (2021b). Financing Clean Energy Transitions in Emerging and Developing Economies 2021, International Energy Agency, Paris. Available Online: https://www.iea.org/reports/financing-clean-energy-transitions-in-emerging-and-developing- economies J. Seel et al., “Interconnection Cost Analysis in the PJM Territory,” Lawrence Berkeley National Laboratory (hereinafter LBNL), January 2023, Available online: https://emp.lbl.gov/publications/interconnection-cost-analysis-pjm. L. Bahiense, G.C. Oliveira, M. Pereira, S. Granville. 2001 A mixed integer disjunctive model for transmission network expansion, IEEE Transactions on Power Systems, vol. 16, pp.560-565. L. Reed, M. Granger Morgan, P. Vaishnav, D. Erian Armanios. 2019. “Converting existing transmission corridors to HVDC is an overlooked option for increasing transmission capacity”, Proc. of National Academy of Science, vol.116 (28), pp. 13879-13884 M. Einberger and C.Teplin.2022. The Best Time to Plan Transmission Was 15 Years Ago. The Second- Best Time Is Now. Rocky Mountain Institute. April. Available online: https://rmi.org/the-best-time-to- plan-transmission-was-15-years-ago-the-second-best-time-is-now/ M. Fürsch, S. Hagspiel, C. Jägemann, S. Nagl, D. Lindenberger, E. Tröster. 2013. “The role of grid extensions in a cost-efficient transformation of the European electricity system until 2050” Applied Energy,vol. 104, pp. 642-652. N. Filatoff. 2021. Synchronous condensers to lift constraints on renewables in South Australia, PV Magazine, August 2021. P. Brown and A. Botterud. 2021. “The Value of Inter-Regional Coordination and Transmission in Decarbonizing the US Electricity System”, vol. 5(1), pp.115-134, January. 24 P.Schell et al. 2013. Using Dynamic Line Rating to minimize curtailment of Wind power connected to rural power networks, Ampacimon (Belgium) Working Paper. Available online: https://www.ampacimon.com/assets/ab1f0112-0341-4920-9999-c181482f2e05/using-dynamic-line- rating.pdf R. Gramlich. 2023. “No energy transition without transmission.” In Harvard Electricity Planning Group Virtual Session, 2023. Available online: https://hepg.hks.harvard.edu/publications/no-energy-transition- without-transmission R. Hemmati, R.-A. Hooshmand, A. Khodabakhshian. 2016. “Coordinated generation and transmission expansion planning in deregulated electricity market considering wind farms, Renewable Energy, vol. 85, pp.620-630. R.P. Singh. 2020. Revisiting the National Renewable Energy Policy, Economic Times, July 11 2020. Available online: https://energy.economictimes.indiatimes.com/energy-speak/opinion-revisiting-the-national- renewable-energy-policy/4380 S. Dehghan, N. Amjady, A. Kazemi. 2014. Two-stage robust generation expansion planning: a mixed integer linear programming model, Power Systems, IEEE Transactions on Power Systems, vol.29, pp. 584-597. S.Pye et al. 2021. Climate finance for grid investments in emerging and developing economies. Available Online: https://climatecompatiblegrowth.com/wp-content/uploads/GGI-Climate-Finance.pdf S. Vega Araujo et al 2023. Locking-in sustainable pathways: just transition, wind energy and local communities in La Guajira, Colombia, Stockholm Environmental Institute. T. Akbari, M. Tavakoli Bina. 2014. A linearized formulation of AC multi-year transmission expansion planning: a mixed-integer linear programming approach, Electrical Power Systems Research, Vol 114, pp. 93-100. Transpower New Zealand. 2012. Scheduling Pricing and Dispatch Software: Model formulation, System Operation Division, Version 6, Wellington, New Zealand. US Department of Energy. 2022. Queued Up but in Need for Transmission: Unleashing the Benefits of Clean Power with Grid Infrastructure, Office of Energy Policy, DOE OP-0015, April 2022. V. Kwaka. 2022. South Africa leading the way toward a low-carbon economy and resilient society, November 22, World Bank blog: https://blogs.worldbank.org/africacan/south-africa-leading-way- toward-low-carbon-economy-and-resilient-society World Bank. 2020. Fifth Power System Development Project, Implementation Completion and Results Report, 2020. Available online: https://documents1.worldbank.org/curated/en/191051598631677294/pdf/India-Fifth-Power-System- Development-Project.pdf 25 World Bank. 2021. Country Climate Change Development Reports, CCDRs are available online: https://www.worldbank.org/en/publication/country-climate-development-reports World Bank. 2022. Financing Energy Transitions in the Power Sector, COP27: Scaling Up to Phase Down: Financing Energy Transitions in the Power Sector, November 2022. Available online: https://live.worldbank.org/events/cop27-catalizing-financing-energy-transition-power-sector World Bank. 2023. Scaling Up to Phase Down: Financing Energy Transitions in the Power Sector, White Paper, April 2023. Y. Zhang et al. 2016. Wind energy rejection in China: Current status, reasons and perspectives, Renewable and Sustainable Energy Review, vol.66, pp.322-344. Z. Huang, I. Smolenova, D. Chattopadhyay, C. Govindarajalu, J. de Witt, T. Remy and I. Deluque Curiel. 2021. ACT on RE+FLEX: Accelerating Coal Transition Through Repurposing Coal Plants into Renewable and Flexibility Centers, IEEE Access, 9: 84811-84127. 26