Modeling Infrastructure Delays and Congestion for Large-Scale Power Systems

Abstract

This paper analyzes the electricity dispatch process and the importance of predicting infrastructure construction delays and network congestion using a system dynamics methodology. Two scenarios of renewable energy integration within Colombia are examined, together with their impact on electrical generation, pricing, and the growth of transmission networks. The results indicate that delays significantly influence the system’s long-term development. Incorporating network congestion into energy dispatch significantly alters investment requirements for generation and transmission, emphasizing network constraints that an idealized dispatch would neglect. Considering congestion highlights the necessity of incorporating delays into energy planning, as these impacts are overlooked in models that fail to account for actual network limitations. This paper emphasizes the strategic significance of considering delays and congestion to attain more realistic and successful planning in extensive power networks.

1. Introduction

Globally, investment in generation projects utilizing non-conventional renewable energy sources has markedly risen. This expansion is propelled by the need to diversify the energy portfolio, reduce CO₂ emissions, and improve the reliability of energy provision [1,2]. To attain the decarbonization objectives of the electricity sector and enable the shift to alternative energy sources, it is imperative to invest in the enhancement and modernization of transmission networks. This investment would facilitate the incorporation of additional renewable generation capacity, alleviate limitations, and improve the efficiency of current power plants [3,4,5].

Factors contributing to the swift global adoption of renewable capacity include the reduced development timelines for unconventional renewable projects compared to traditional power plants. The augmentation of generation capacity and transmission infrastructure is significantly contingent upon synchronizing their planning and development timelines. On the one hand, if generation expands more rapidly than transmission, coordination may become progressively intricate. Therefore, both capacities must be cultivated together [3,6,7,8]; on the other hand, transmission infrastructure, in turn, has encountered various barriers that prevent it from expanding at the same rate as generation plants. Although the intensity of these barriers may differ from country to country, the challenges are broadly similar in most cases [6,9].

Regarding transmission capacity and network congestion, conventional long-term power system models frequently employ a “copperplate” assumption, which essentially refers to the grid as possessing infinite capacity, to streamline dispatch and expansion decisions [10,11]. Under this idealized premise without constraints, generation can be allocated with no restrictions to satisfy demand universally. Nonetheless, if the integration of renewables increases and power distribution becomes more geographically uneven, this simplification may yield deceptive outcomes. Recent studies have underscored the necessity of explicitly modeling network congestion and transmission constraints for enhanced accuracy in expansion planning. Incorporating realistic grid constraints in capacity expansion models prevents impractical buildouts that would be desirable just within an unconstrained network [12,13,14]. Neglecting transmission constraints in power expansion planning can result in inferior investment choices [11]. An NREL survey indicated that most utility planning models fail to co-optimize production and transmission, relying instead on rudimentary “pipe” or copperplate representations; this absence of co-optimization may result in suboptimal and elevated-cost solutions, particularly in scenarios with significant renewable energy integration. In conclusion, integrating congestion into modeling prevents the overestimation of an optimal dispatch and guarantees that growth plans remain viable under actual network conditions [15].

Not considering network constraints might cause efficiency deficits and operational expenses that arise in actual systems. When grid limitations are restrictive, power cannot transfer from low-cost generators to load centers, resulting in price discrepancies across areas and the reliance on more costly local generation [16,17]. This inefficiency has been quantified: a study of European power markets indicated that straying from an optimal, fully network-aware dispatch (nodal pricing as the first-best benchmark) results in welfare losses of approximately 4.6% [18]. Restrictions on otherwise accessible wind or solar energy output due to transmission constraints are anticipated to increase with further integration of variable renewable sources [19,20]. This not only misuses carbon-free energy, but also results in elevated fuel expenses as fossil units are deployed instead of the curtailed renewables.

Given the essential function of transmission networks in supporting a growing and diverse energy mix, modeling and assessing grid growth and network congestion are crucial for a thorough comprehension. Disregarding congestion restrictions may result in erroneous dispatch decisions, miscalculations of renewable integration capabilities, and ineffective investment plans [21,22]. By explicitly integrating realistic transmission limits and network bottlenecks, planners and policymakers can more accurately evaluate the actual costs, benefits, and risks associated with generation expansion. Ultimately, these methodologies guarantee more dependable, efficient, and sustainable power systems that are consistent with both present and future decarbonization objectives [23,24].

In this paper, system dynamics (SD) is utilized to model and examine electricity market situations, comparing a model that explicitly incorporates network congestion with one that excludes it. We seek to emphasize the significant impact of transmission constraints on facilitating renewable adoption and decreasing CO₂ emissions. These scenario-based simulations will find effective solutions to expedite the adoption of renewable energy sources and guide targeted policy actions to reduce infrastructure expenditures. This approach aims to align with Colombia’s national objective of a 30% emissions reduction, providing policymakers and stakeholders with practical insights on how congestion-aware planning can enhance system reliability, and further long-term sustainability goals [25].

Several existing modeling platforms have been created to aid large-scale energy planning, such as NREL’s ReEDS model in the United States and the scenario-based assessments from ENTSO-E and the Ten-Year Network Development Plan (TYNDP) in Europe [26]. High spatial and temporal resolution, as well as thorough technoeconomic assumptions spanning several regions and technologies, define these models. However, they are not meant to capture behavioral feedback, investment inertia, or long-term system dynamics and usually run in a linear optimization or partial equilibrium framework. By explicitly including transmission congestion and simulating investment responses to changing conditions, our system dynamics approach emphasizes the co-evolution of generation and transmission capacity over time [27]. Especially in situations where data availability or computational complexity may limit the use of large-scale optimization models, this paradigm enables the investigation of emergent behavior under uncertainty and policy interventions.

This paper analyzes long-term power market behavior under various infrastructure assumptions using two scenarios inside an SD framework. The first scenario assumes idealized transmission conditions free from network congestion, enabling unhindered flow of electricity between areas. The second scenario incorporates reasonable limitations on grid capacity and transmission congestion. Through a comparison of different methods, the model demonstrates how neglecting grid limitations could result in overly optimistic estimates of installed capacity, power pricing, and connectivity requirements. The outcomes highlight the major influence of transmission restrictions on investment decisions and system development, as well as the necessity of incorporating such limits into national energy planning procedures.

This paper is structured as follows. Section 2 outlines the theoretical and policy adjustments for the integration of renewable energy sources and the reduction of network congestion. Section 3 outlines the materials and methods utilized. Section 4 delineates the simulation findings obtained from the modeling methodology and elucidates the results of the simulation model. Section 5 discusses the implications of the findings. Section 6 ultimately delineates the conclusions and suggests potential avenues for future research.

2. Perspectives on Coordinated Generation and Transmission Planning with Grid Limitations

Models for the development of electricity markets offer a necessary understanding for the formulation of policies and wise decision-making. These models can be very helpful in assessing ideas meant to improve the electrical system or in spotting possible situations resulting from the introduction, change, or elimination of some conditions. Various scenarios about the inclusion of non-conventional renewable energy sources, energy storage devices, distributed generation, and possible changes in demand patterns within the system found by different authors are presented below [28]. Most of these authors emphasize the need for coordination among the components of generation and transmission in planning and development [29].

First, various obstacles have been experienced by renewable energy sources trying to enter the market. Although possible remedies have been found, every nation’s electrical system is unique; hence, the strategies carried out could differ [30,31]. This emphasizes the need to complete customized research for every market, evaluating incentives for fresh investments and looking at the effects that delays could have on system dependability. Furthermore, as they can cause delays in the start of the infrastructure construction for transmission and generation, environmental and social elements progressively become very important for project feasibility [32]. Some models within the generation component posit that investors augment capacity by optimizing costs, resource availability, risk, and environmental implications, guided by current incentives and profitability assessments to inform their investment decisions [33,34,35,36]. Conversely, the central planner must evaluate generator patterns and direct the installation of additional capacity through incentives and rules to ensure a sufficient degree of reliability [2,32,37].

In the transmission component, authors examine the function of a central planner who seeks to optimize societal welfare by enabling more economical energy transfer between regions and minimizing congestion expenses [38]. Nonetheless, the emergence of new technologies, such as batteries and distributed resources, provides various ways to accomplish this objective. However, insufficient transmission capacity hinders the integration of new renewable energy facilities, creating uncertainty and deterring investment. Consequently, synchronizing transmission and generation expansion is crucial for sector development [39,40].

Regarding network congestion, ref. [41] examines its alleviation through demand-side strategies and dispersed resources. It introduces an efficient scheduling approach that synchronizes demand response programs (DRPs) with the integration of distributed generation (wind units), employing DC load flow and transmission distribution parameters. By strategically timing demand response actions and positioning distributed generation units, the approach mitigates transmission overloads and substantially enhances the available transfer capability (ATC). Alternatively, ref. [42] investigates the utilization of battery energy storage systems (BESS) as a centrally managed flexibility resource to alleviate congestion in distribution networks. The BESS can enhance grid efficiency and decrease the need for infrastructure upgrades by deliberately collecting surplus energy during off-peak periods and supplying electricity during peak times, thereby mitigating feeder overloads and voltage problems. The strategies are evaluated using an actual Spanish distribution network over a one-week period. The research elucidates the optimization of battery location and dispatch for congestion control in a centralized manner through advanced network modeling.

In [43], a bi-level congestion management technique that integrates the transmission system operator (TSO) and the distribution system operator (DSO) is presented. The TSO mitigates congestion costs at the upper level by re-dispatching generation and modifying nodal prices within the transmission grid; concurrently, the DSO calculates dynamic compensation prices for flexible loads in the distribution network at the lower level to encourage demand response that bolsters the transmission system. An optimal congestion mitigation plan is attained through iterative interaction between the two tiers, ensuring that transmission and distribution resources collaborate to alleviate network overloads. Results: the simulation demonstrates that the coordinated TSO–DSO strategy can more effectively eradicate transmission bottlenecks compared to solely depending on conventional generator re-dispatch. The authors of the case study compare three programs: (a) multiple adjustments (traditional re-dispatch), (b) fixed dispatch plan, and (c) the proposed market-based coordination.

In [44], an SD model to examine the impact of introducing financial transmission rights (FTRs) on transmission grid congestion and expansion decisions is developed. The authors simulate prolonged interactions involving load increase, transmission line revenues, network congestion, and investment in additional lines within market circumstances, utilizing FTR-based congestion management. Sensitivity analyses, which involve altering generator bid prices, expansion costs, and FTR market maturity, demonstrate that integrating FTRs with expansion investments can render grid improvements economically feasible and market-driven. The study indicates that FTRs offer a market mechanism to alleviate congestion and indicate where transmission development is most beneficial, assisting centralized planners and operators in their decision-making processes.

In [45], a system dynamics simulation is utilized to investigate the effects of policy-driven adoption of distributed solar photovoltaics (PVs) and electric vehicles (EVs) on the electric grid, including potential congestion challenges. The model integrates feedback mechanisms among governmental incentives, infrastructure advancement, consumer adoption rates, and grid demand. The emphasis is on central authorities (policymakers), as the findings highlight the significance of supportive rules and coordinated planning in facilitating the transition. The Colombian case study indicates that with favorable policies, significant adoption of PV systems and EVs is attainable, resulting in decreased CO₂ emissions and a transition of the vehicle fleet to electric. It also underscores the necessity to modify grid capacity and operations to prevent new congestion issues.

In [46], centralized congestion management strategies in the provincial power spot markets of China are analyzed. This paper evaluates the current methodologies, including both direct operator interventions and market-oriented approaches, while proposing enhanced strategies in three areas: (a) cost optimization through generation rescheduling and adjustments to the available transfer capability (ATC), (b) establishment of a “congestion cost pool” for surplus redistribution and price signaling, and (c) implementation of a two-tier transmission rights trading market to hedge against and alleviate congestion risk. The authors illustrate through a provincial case study that their integrated strategy, which merges dispatch adjustments with financial rights trading, substantially reduces congestion resolution costs and enhances system operator efficiency, achieving improvements of approximately 62% in simulations. The results provide guidance to grid operators and regulators on comprehensive, centrally coordinated strategies for managing transmission overload within a market context.

Particularly considering the complicated interdependencies among several stakeholders, a comprehensive knowledge of the electrical system is essential. SD provides a robust approach for creating integrated models that incorporate the co-evolution of generation and transmission capacity, therefore enabling the forecasting and assessment of future expansion scenarios. This includes evaluating emerging technologies, demand variability, infrastructure investment delays, and legislative impacts. By allowing flexible, data-informed decision-making in an ever-changing energy environment, such an approach strengthens policy formulation. Larsen and Bunn [47] used SD to investigate market behavior, capacity investment cycles, and strategic risks in deregulated electrical systems; they did not combine the simultaneous modeling of transmission congestion with coordinated capacity expansion. To the best of our knowledge, no study to date has combined network congestion analysis with concurrent planning of both generation and transmission infrastructure in a long-term growth context. Through a system dynamics model that explicitly simulates congestion effects, investment trade-offs, and their impact on infrastructure development paths, our work closes that gap. Consequently, it is important to develop a new SD model to investigate and assess prospective system expansion possibilities during the current energy transition. This model will consider delays in generation and transmission projects, evaluate different expansion strategies, and examine the importance of explicitly simulating network congestion compared to excluding it.

The following section presents the case study and the simulation model, detailing the data sources, the dynamic hypothesis, and the underlying model structure.

3. Comprehensive Case Study and Simulation Framework for Colombia’s Power Grid

3.1. Colombian Electricity Market

Colombia was chosen as the principal case study because of its advantageous position in the equatorial region of South America, which offers ample and rather uniform solar irradiation throughout the year. The country possesses substantial potential for the implementation of PVs and wind technology, with an average solar radiation of ca. 4.5 kWh/m²/day [48]. Furthermore, most of Colombia’s wind and solar resources are situated in its northern areas, requiring an analysis of energy flow patterns and the planning of adequate transmission infrastructure to efficiently integrate these renewable sources [48]. This analysis underscores the necessity of expanding generation and transmission capacity to alleviate congestion issues and prevent increased electricity prices, providing a more dependable and economical power system.

Colombia’s generation and transmission network is divided into five main regions, as shown in Figure 1, along with their respective departments—Caribe, Antioquia, Nordeste, Oriental, and Suroccidental. In each region, a single aggregated hydro dam represents the area’s total hydro capacity.

The opportunity cost of water depends on the expected level of the reservoir and the offer price of dispatchable generation technologies [49].

Table 1. Opportunity cost of water.

Level of Reservoir (%)	Opportunity Cost
0.2	Scarcity price
0.5	Most expensive thermal price
0.8	Least expensive thermal price
1.0	Marginal cost

indicates the values for the opportunity cost of water considered in the model.

It is important to clarify the meaning of the parameter “most expensive thermal price”. This value corresponds to the marginal cost of the most expensive thermal technology included in the country’s generation mix. In the context of the model, it serves as the reference price for the opportunity cost associated with hydroelectric generation and storage decisions. Specifically, this cost acts as a threshold beyond which hydroelectric reserves are dispatched, ensuring their optimal use during periods of high demand or limited supply from cheaper sources. It prevents the reservoir from being dispatched when its level is low and alternative generation technologies are available. This adjustment helps regulate the reservoir’s level, optimizing water use for higher-profit periods and ensuring long-term generation efficiency. Further methodological details can be found in [49].

Additionally, the initial generation capacities for each technology in the five regions, as well as the distributed energy resource (DER) capacities, were updated with data from the system operator XM, reflecting the system’s status as of January 2025 [50,51].

Table 2. Initial generation and DER capacities [50,51].

Region	DER Capacity [MW]	Technology	Capacity [MW]
Caribe	42.40	Biomass 1	2.0
		Solar	1.18
		Hydro dam	338.0
		Thermal 1 2	709.0
		Thermal 2 3	3088.0
Antioquia	49.73	Biomass 1	5.0
		Solar	3.0
		Run-of-the-river	270.0
		Hydro dam	5707.0
		Thermal 1 2	4.0
		Thermal 2 3	433.0
Nordeste	43.10	Solar	49.0
		Hydro dam	1901.0
		Thermal 1 2	693.0
		Thermal 2 3	714.0
Oriental	28.15	Biomass 1	23.0
		Solar	184.0
		Run-of-the-river	185.0
		Hydro dam	2006.0
		Thermal 1 2	226.0
		Thermal 2 3	2.0
Suroccidental	68.48	Biomass 1	167.0
		Solar	514.0
		Run-of-the-river	361.0
		Hydro dam	2450.0
		Thermal 1 2	27.0
		Thermal 2 3	529.0

Note: ¹ bagasse and biogas; ² coal and diesel; ³ natural gas, jet A-1 and fuel oil.

summarizes the values used.

3.2. Model Description

An SD model was developed to estimate the expansion needs of a power system in the long term, specifically the generation and transmission components, including non-conventional renewable energy, DERs, and battery energy storage systems (BESS).

Figure 2 presents the causal loop for the expansion of the electricity market. This causal loop diagram illustrates the dynamic interactions inside an electricity system, emphasizing the interdependence among investments, prices, and infrastructure development. A critical component of the system is grid congestion, which arises when transmission capacity is inadequate to transport electricity from generation sources to demand centers. As congestion escalates (Loop R2), it elevates spot prices and constraint costs (Loop B2), which directly influence the total electricity bill incurred by consumers. The increased expenses may prompt investments in both generation and transmission capacity (Loops B8 and B9) to alleviate congestion. Nevertheless, until those investments are realized, congestion endures and may exacerbate if supply escalates more rapidly than infrastructure enhancements.

Restriction costs denote the penalties or inefficiencies incurred from rerouting or limiting power flow due to congested transmission lines. These expenses result directly from grid congestion and significantly contribute to increasing electricity tariffs (Loop B2), imposing economic strain on both suppliers and customers. The system seeks to mitigate congestion and associated costs by investing in transmission infrastructure (Loop R4) and distributed energy resources such as photovoltaic panels (Loops R1 and B4). This dynamic underlines the need for proactive grid planning and the use of decentralized energy in alleviating congestion-related inefficiencies and stabilizing energy markets.

In the generation module, projects are selected based on a desired system margin, and then enter the licensing and construction phases, passing through their development stages until they reach readiness to operate. Depending on the stage of each project, an estimated completion date is assigned, and its projected capacity is calculated over planning horizons of 3, 5, 7, and 9 years. A project’s capacity becomes effectively installed only if the transmission network has sufficient margin to accommodate the new addition. In Appendix A, hourly solar and wind capacity factors are shown.

In the transmission module, projects are triggered by demand growth, interconnection needs, and generation expansion. However, only those related to interconnections are subject to delays. These delays span from the development of the expansion plan and public bidding process to licensing and construction, resulting in the deployment of transmission lines. As shown in

Table 3. Transmission project timeline.

Stage	Time [Months]
Expansion plan development	12
Public tender	21
Investor selection	4
Licensing process	42
Construction process	35
Total	114

, the total duration for these stages is 114 months. As in the generation module, investment decisions in transmission are also based on future system projections.

To perform an economic dispatch integrating the SD model and the system’s congestion, Vensim was connected to Python (3.11.7) using Vensim’s Dynamic Link Library (DLL). This setup enabled Python to load and run the Vensim model, set simulation parameters (such as region, active technologies, and forecast horizons), and exchange variable values with PyPSA (Python for Power System Analysis) developed by [52].

At each simulation step, the script retrieved data from Vensim, including demand, availability factors, projected generation capacity, and offer prices, and used this information to construct the corresponding PyPSA network model. Each region was represented as a node, with interregional connections modeled as transmission lines and technologies assigned accordingly.

Two dispatch simulations were performed in PyPSA for each time step. The first assumed unlimited transmission capacity to estimate ideal dispatch and identify potential interconnection needs. Based on this, Vensim determined infrastructure upgrades, and the new transmission limits were sent back to PyPSA for a second, congestion-constrained dispatch. The resulting interconnection capacities, dispatches, and system prices were written back into Vensim, allowing dynamic feedback into the long-term SD model.

The overall workflow is illustrated in Figure 3. It shows the role of DLL functions in managing the communication between Python and Vensim, with dotted lines indicating function arguments. Dotted lines ending in square brackets represent the subscript structure of variables, such as year, hour, technology, and zone, exchanged between platforms throughout the simulation.

The following section presents the modeling scenarios developed for the proposed green hydrogen production facility, along with the corresponding results and analysis derived from the simulations.

4. Results

This section presents a comprehensive comparison between the two scenarios, providing the quantitative evidence needed to support policy formulation. It not only confirms theoretical expectations, but also highlights the extent of differences resulting from including or excluding transmission constraints. These measurable differences help decision-makers assess the benefits of realistic congestion modeling and the risks associated with relying on oversimplified planning approaches.

To evaluate the importance and comprehensively grasp the potential constraints of utilizing an optimal dispatch model in electricity system planning, two separate dispatch modeling scenarios were analyzed. The first scenario was formulated under the assumption of optimal, congestion-free distribution throughout the nation, ignoring any network constraints. In contrast, the second scenario clearly included limits related to network congestion in electricity transmission throughout the five investigated locations. This work aimed to elucidate significant differences between the two modeling approaches, emphasizing the essential role of integrating transmission restrictions and grid congestion factors in the formulation of thorough and precise long-term energy planning models.

Figure 4 compares the installed energy capacity and peak electricity demand predictions for Colombia, based on two distinct modeling scenarios from 2025 to 2050.

Figure 4a illustrates the situation modeled without accounting for network congestion limits. In this optimized dispatch scenario, the energy system primarily depends on thermal power plants (Thermal 1 and Thermal 2), hydroelectric dams, and geothermal resources. A gradual rise in renewable generation capacity is noted, especially from DERs, PVs, and both onshore and offshore wind technologies. By 2050, the total installed capacity is projected to reach roughly 60 GW. This scenario demonstrates considerable advancement in renewable deployment. Yet, the lack of network constraints leads to an overestimation of the overall capacity requirements relative to more realistic models. Peak demand increases steadily over the planning horizon and stays beneath the total installed capacity, ensuring adequate reserve margins to address transmission-related limitations.

Conversely, Figure 4b depicts the situation in which network congestion limitations among the five examined locations are explicitly incorporated. This structure necessitates less installed capacity to satisfy regional demands while adhering to transmission constraints. Renewable energy sources, particularly PVs and DERs, demonstrate significant growth, with the total installed capacity nearing 50 GW by 2050. This augmentation reinforces the imperative for enhanced capacity to mitigate regional transmission constraints and guarantee a dependable supply. Peak demand exhibits an upward trend and remains well below the total capacity, although with a reduced reserve margin, which may reflect the system’s flexibility requirements.

Though the unconstrained model, Figure 4a, allows power flow unhindered between areas, some restrictions still limit its capacity to fully meet demand. Under this situation, solar generation increases in line with the yearly growth rate of 2.5% for the demand in power. However, because of their capacity factors—25% for solar PVs and 30% for wind—which reduce their effective energy output, solar and wind technologies are not very plentiful. The model also runs one typical day each month, limiting the system’s ability to dynamically balance supply and demand over shorter timescales or account for seasonal variations. Therefore, even with increasing installed capacity, the system could still have times when demand is unsatisfied, especially in cases when backup or dispatchable resources are insufficient. In contrast, the congestion-aware model, Figure 4b, requires each area to meet its own demand depending on a local merit order within transmission capacity limits. This reflects the spatial difficulties of grid extension under physical limits and produces a more reasonable distribution of generating resources.

Collectively, these data highlight the substantial influence of integrating transmission congestion restrictions in long-term power system planning. The comparison illustrates that idealized models may overestimate the necessity for supplementary capacity, whereas congestion-aware models offer a more precise depiction of the investments and infrastructure essential for facilitating a dependable and sustainable energy transition.

Figure 5 depicts daily electricity generation by technology under two different predicted scenarios in Colombia from 2025 to 2050. Figure 5a illustrates the scenario disregarding network congestion limitations. In this optimal dispatch scenario, thermal power plants (Thermal 1 and Thermal 2) initially play a crucial role in the daily electricity supply, offering essential flexibility. Hydroelectric dams continue to serve as a reliable and significant resource for the whole planning period. Nonetheless, the output of renewable energy, especially from DERs, PVs, and wind technologies (both onshore and offshore), is anticipated to increase, leading to the total daily generation surpassing 400 GWh/day by 2050. Despite substantial advancements in renewable energy, the idealized scenario fails to account for system complexity, potentially neglecting instances when transmission constraints may impede efficient dispatch, hence diminishing the necessity for supplementary reserve capacity.

Figure 5b fully depicts the situation that includes network congestion limits. In this context, thermal power generation plays a more significant and enduring role initially, highlighting the necessity to efficiently manage network constraints. Hydroelectric generation consistently offers a reliable baseline. The input from renewable sources, especially distributed energy resources, PVs, and wind energy, is significantly elevated, approaching 500 GWh/day by 2050. The augmented renewable generation relative to the optimal dispatch scenario indicates the need for enhanced energy production capacity to mitigate transmission constraints and guarantee the dependability and resilience of the electricity supply. This model exhibits increased variability and elevated peaks in renewable production, underscoring the necessity for sophisticated planning and investments in grid flexibility to manage these swings.

In summary, the comparison of both scenarios highlights the necessity of incorporating congestion restrictions in long-term energy modeling. The congestion-aware scenario more precisely reflects the necessary investments and generation capacity, underscoring potential deficiencies in simplistic, congestion-free models.

Figure 6a,b depicts the development of transmission in Colombia by type from 2025 to 2050, contrasting the scenarios modeled without and with network congestion limits, respectively.

In Figure 6a, representing the scenario devoid of network congestion limitations, the transmission infrastructure undergoes a moderate yet consistent increase over the simulated period. Transmission lines associated with generation and demand primarily constitute most of the overall network expansion. By 2050, the entire length of transmission lines is projected to be roughly 36,000 km, with the majority allocated for generation. The growth of interconnection transmission infrastructure is comparatively limited, implying a diminished necessity for improved connectivity between regions in this optimal dispatch scenario.

In contrast, Figure 6b illustrates the situation with network congestion limits explicitly accounted for. A significant expansion of transmission is noted, with the overall line length just over 37,000 km by 2050. Compared to the previous scenario, the transmission lines needed from regional interconnection are ca. 1000 km longer, underscoring the imperative for strong interregional connectivity to efficiently address transmission bottlenecks. The transmission for demand and generation is lower than in the unconstrained scenario, 144 km and 67 km, respectively, highlighting the critical function of transmission infrastructure in integrating geographically scattered renewable generation resources and reliably satisfying regional power demand.

Analyzing these scenarios underlines the necessity of clearly estimating transmission congestion. Disregarding network restrictions may result in an underestimation of transmission infrastructure needs, especially for interregional connectivity, potentially causing insufficient investments and operational weaknesses in long-term energy planning.

Figure 7 illustrates a comparative analysis of electricity costs in two unique scenarios: with and without network congestion constraints, spanning from 2025 to 2050. The vertical axis denotes the total electricity tariff in USD/MWh, while the horizontal axis encompasses the simulation timeline. The image distinctly demonstrates the influence of network congestion on power pricing trends.

In the context of congestion restrictions (orange line), energy tariffs remain elevated throughout the examined timeframe, particularly in the initial years (2025–2035), when the system relies more heavily on flexible thermal generation due to ongoing infrastructure development. In this scenario, tariff peaks surpass 500 USD/MWh in the early years, indicating the financial consequences of restricted transmission capacity and the failure to efficiently allocate lower-cost renewable energy among areas. As transmission infrastructure advances and renewable energy integration rises, tariffs progressively stabilize, reducing the disparity between the two scenarios.

Conversely, the congestion-free scenario (blue line) exhibits reduced and more consistent tariff values, signifying that optimal conditions with unrestricted energy dispatch result in a more efficient utilization of generation resources and diminished system costs. Such assumptions, however, neglect the practical constraints of electrical networks, potentially underappreciating the actual cost determinants in long-term planning.

Figure 7 underscores the significance of incorporating transmission restrictions in tariff modeling. Disregarding these limits may lead to underestimated tariffs and an inaccurate portrayal of the actual investment necessary for a cost-efficient and dependable renewable energy transition.

Finally,

Table 4. Comparison between the scenarios with and without network congestion.

	2025		2050
	Without Congestions	Congestions	Without Congestions	Congestions
Hydro capacity [MW]	13,218.06	13,218.06	16,002.28	15,647.43
Solar capacity [MW]	1926.32	1926.84	18,574.80	16,534.00
Wind capacity [MW]	0.00	0.00	9628.20	9883.56
Thermal capacity [MW]	6426.19	6426.19	7138.70	7127.70
DER capacity [MW]	230.86	230.86	6599.01	832.12
Share of renewables [%]	70.48	70.48	86.95	86.42
Transmission for demand [km]	5116.19	5116.19	9375.54	9231.87
Transmission for generation [km]	8707.38	8707.38	21,882.10	21,815.1
Transmission for interconnection [km]	3860.70	3860.70	4888.60	6030.4
Total transmission [km]	17,684.30	17,684.30	36,146.20	37,077.4
Availability margin [%]	51.62	53.90	46.59	46.24
Electricity tariff [USD/MWh]	257.01	370.77	108.54	98.23

presents a comparative analysis of both scenarios, specifying the first and final values for centralized generation capacity, solar and wind capacity, the proportion of renewable energy, the total transmission length by component, the availability margin, and the overall electricity tariff. Both scenarios commence with equivalent installed capacities: hydroelectric (13,218.06 MW), solar (ca. 1926 MW), thermal (6426.19 MW), and DER generation (230.86 MW), with no wind capacity present. Significant differences arise by the conclusion of the study period; in the absence of congestion limitations, solar capacity attains a superior level of 18,574.80 MW compared to the congestion scenario of 16,534 MW; and wind capacity is marginally reduced at 9628.20 MW in contrast to the congestion constraints scenario of 9883.56 MW. DER capacity markedly increases to 6599.01 MW in the uncongested scenario, in stark contrast to merely 832.12 MW under congested conditions. The share of renewable energy in the unconstrained scenario, 86.95%, barely surpasses the 86.42% in the constrained scenario. The transmission infrastructure, particularly interconnection lines, is more extensive under congestion limitations (6030.4 km against 4888.60 km), leading to a greater overall transmission length (37,077.4 km compared to 36,146.20 km). The scenario involving congestion constraints leads to higher initial electricity tariffs (370.77 USD/MWh versus 257.01 USD/MWh) but results in lower tariffs by the conclusion of the modeling period (98.23 USD/MWh compared to 108.54 USD/MWh), demonstrating the long-term advantages of thorough infrastructure planning and congestion management.

Table 4 compares the key infrastructure indicators across both scenarios, quantifying the influence of these assumptions. Policymakers aiming to develop effective long-term planning systems and investment plans will find these insights particularly relevant.

5. Discussion

The model results demonstrate the importance of considering transmission constraints to evaluate the potential evolution of electricity systems in the long term, as they have a significant impact on generation capacity requirements, prices and investment decisions.

Table 4 highlights the policy relevance of comparing the scenarios with and without congestion, as it reveals significant differences in infrastructure outcomes and economic impacts. For instance, the congestion-aware model estimates a total installed capacity of just 50 GW—a 17% reduction—compared to the unconstrained scenario which reaches 60 GW by 2025. This difference suggests that neglecting transmission congestion leads to overinvestment, potentially resulting in stranded assets or inefficient capital allocation. Policymakers should therefore ensure that real transmission constraints are considered in expansion plans to avoid overbuilding. Furthermore, under the congestion scenario, daily renewable generation reaches up to 500 GWh/day, compared to about 400 GWh/day in the unconstrained model (Figure 5). This demonstrates how the system can strategically position additional renewable resources where they are most effective in reducing congestion and balancing supply, rather than solely in areas with favorable natural conditions but limited transmission capacity.

The scenario with congestion shows tariff peaks above 500 USD/MWh, while the unconstrained scenario remains below 300 USD/MWh. This further emphasizes the importance of realistic modeling (see Figure 7). These findings support dynamic tariff changes and congestion pricing schemes as tools for demand-side control and investment signals. Reflecting the systematic need to invest in transmission to unlock access to lower-cost renewable energy, the total length of interconnection lines increases by 1141.8 km in the congestion-constrained scenario (Table 4). This suggests that targeted efforts in interregional connectivity can reduce reliance on more expensive, dispersed generating sources. These data-driven insights advocate for integrated planning approaches, where generation incentives are directly linked to transmission capability and regulatory bodies incorporate congestion simulations into long-term planning processes. Such synergy can accelerate Colombia’s change toward a more flexible, resilient, and cost-effective energy system.

These results further emphasize the importance of considering grid limitations when determining capacity requirements within the context of renewable energy growth in Colombia. Assuming ideal transmission conditions could lead to overestimating installed capacity requirements, which impacts investment plans and policy decisions. By incorporating spatial distribution restrictions and congestion, a more realistic approach yields a technically feasible and economically sound system design. This underscores the need for integrated planning instruments that specifically account for transmission constraints alongside renewable energy sources.

Establishing reasonable energy prices and tariffs relies on accurate modeling of transmission congestion, which is crucial for promoting investment in generation capacity. The results show that ignoring congestion restrictions can lead to significantly inflated tariffs, potentially distorting policy development and investment decisions. Tariffs influenced by network congestion more accurately reflect actual grid conditions and the costs associated with integrating renewable energy. Therefore, realistic tariff signals are essential to encourage timely and adequate investment in the infrastructure supporting renewable energy generation and transmission. To provide correct economic signals, build investor confidence, and enable a strong, economically sustainable energy transition, policymakers must explicitly account for transmission constraints.

Figure 6 presents a comparative study of power prices, emphasizing the urgent need for coordinated planning of transmission and renewable production capacity expansion. Ignoring network congestion in the design process leads to overly optimistic electricity cost estimates, which can mislead investment decisions. Aligning transmission system development with the rate of renewable capacity expansion must be a top priority for policymakers. Supported by robust regulatory frameworks, implementing policy instruments, such as mandatory coordinated planning between utilities and generation investors, can help alleviate congestion impacts, reduce electricity costs, and provide stronger investment signals to stakeholders.

The higher electricity prices observed during network congestion accentuate the need for specific laws to mitigate the financial impact of transmission congestion. To encourage the strategic placement of generation plants closer to demand centers, policymakers should consider financial mechanisms such as congestion charges, transmission capacity auctions, or tariff-based incentives. Additionally, funds collected through congestion pricing could be allocated to support grid modernization initiatives and expand transmission infrastructure. By easing the financial burden of network congestion and encouraging private-sector involvement in grid projects, these targeted financial tools help Colombia accelerate its transition to a sustainable and economically viable renewable energy system.

6. Conclusions

This research emphasizes that the precise modeling of network congestion limitations substantially affects the anticipated installed capacity of renewable energy resources. In scenarios accounting for transmission congestion, Colombia’s power system requires less installed capacity—nearing 50 GW by 2050—relative to the 60 GW when congestion is disregarded. Consequently, disregarding precise network flows and congestion restrictions may lead to an overestimation of the requisite investments in renewable generation capacity andincreasethe risks of failing to meet long-term electricity demand, alongwith higher costs of power generation.

The findings mark the fundamental importance of DERs, PVs, and both onshore and offshore wind power in meeting Colombia’s renewable energy objectives. Notwithstanding network limitations, these sources exhibit continuous expansion and emerge as the primary contributors to the energy mix by 2050, highlighting their significance in the nation’s long-term sustainability policy.

Transmission expansion scenarios indicate that specifically simulating network congestion results in an increased investment in transmission infrastructure—particularly interregional interconnections—by about 1000 extra kilometers by 2050. Therefore, optimal synchronization between renewable energy generation and transmission system growth is essential for the reliable integration of geographically scattered renewable resources.

Daily energy generation results reveal increased renewable energy integration and heightened unpredictability in renewable outputs when congestion restrictions are explicitly accounted for. The simulated scenario with congestion constraints projects approximately 500 GWh/day of renewable electricity generation by 2050, in contrast to roughly 400 GWh/day in the unconstrained scenario, highlighting the necessity of precise modeling of transmission limitations for a realistic evaluation of grid flexibility and planning requirements.

The incorporation of transmission congestion underscores the necessity of sustaining thermal and hydroelectric generation as flexible resources, especially in the early phases (2025–2035). Thermal generation has a falling trend over time, highlighting the viability and significance of a strategically planned transition to renewable energy that gradually diminishes dependence on the fossil fuel technology.

The results demonstrate that network congestion has a significant impact on electricity prices, particularly in the short-to-medium term. When congestion constraints are considered, electricity tariffs are consistently higher—exceeding 500 USD/MWh in peak periods—due to the limited ability to dispatch lower-cost renewable energy across regions. Over time, as transmission infrastructure is expanded and renewable generation increases, tariff volatility and price differentials between the two scenarios diminish. This highlights the importance of coordinated investment in both generation and transmission infrastructure to minimize electricity costs and ensure a stable and efficient energy transition.

This study highlights the strategic significance of interregional connections and the construction of transmission networks to efficiently balance power supply and demand throughout Colombian regions. Integrating the extensive solar and wind energy potential in the northern region with the substantial hydroelectric resources in the central region necessitates proactive policy and infrastructure development to optimize renewable electricity dispatch and ensure system reliability and environmental sustainability.