Over the past few years, there has been an exponential increase in the demand for all-round reliable telecom connectivity, driven by the accelerated pace of digital transformation across industries and the implementation of next-gen technologies such as the internet of things (IoT) and cloud computing. Moreover, increased reliance on remote work, online learning, e-commerce and e-payments, along with the rise in video calling, messaging apps and social networking sites during and after the Covid-19 pandemic has further heightened the need to ensure uninterrupted communication at all times. To meet the increasing need for seamless communication, telecom infrastructure firms and other players in the information and communication technology (ICT) sector have bolstered their network infrastructure across the country. As the demand for reliable and high-speed connectivity continues to rise, ICT companies are looking to further strengthen their digital communications infrastructure.
The expansion of telecom infrastructure, however, is highly energy-intensive. The construction and operation of telecom towers, continuous power consumption of network equipment, establishment and maintenance of data centres and deployment of optical fibre networks, satellite communications and undersea cables necessitate substantial power usage. Moreover, regular upgrade and maintenance requirements further add to the energy footprint of telecom infrastructure.
Given the significance of telecom networks in digital transformation and their ecological impact, addressing the associated energy consumption becomes a crucial aspect of sustainable infrastructure development. While go-green initiatives have gained traction in the telecom industry in recent years, financial constraints faced by operators have hindered significant investments in establishing the necessary infrastructure for alternative energy generation. Therefore, there is a need to explore cost-effective solutions that help lower the energy quotient of the telecom industry without compromising on service quality.
A look at the energy consumption across various ICT sectors and strategies to mitigate their environmental impact…
Energy footprint of different ICT segments
- Data centres and cloud computing: A large number of industries have started relying on data centres for their data storage and computation requirements. Data centres also serve as the backbone for cloud computing by providing the necessary infrastructure, scalability, reliability, security and cost efficiency to deliver a wide range of computing services over the internet. However, data centres, by design, are energy-guzzling facilities as they require huge amounts of power to process and store data as well as to cool off server racks of computing equipment. Estimates suggest that data centres contribute to approximately 5 per cent of global greenhouse gas emissions. Within a data centre facility, the most substantial share of direct electricity consumption (86 per cent) is attributed to servers and cooling systems, followed by storage drives (11 per cent) and network devices (3 per cent). Data centre servers also generate heat, necessitating additional energy for cooling.
- Tower sites: The lack of consistent grid power at most tower sites has compelled tower companies to resort to non-grid sources, primarily diesel generators. The use of diesel, however, has considerable ecological consequences. The annual carbon dioxide emissions resulting from diesel usage at telecom tower sites are estimated to be 10 million tonnes. This places the telecom industry’s contribution to the country’s total carbon dioxide emissions at around 1 per cent, surpassing the global standard of 0.7 per cent.
- 5G: Despite the potential for 5G to deliver considerably higher speeds than 4G, concerns have emerged about a corresponding surge in energy consumption, assuming constant energy efficiency. Experts attribute this potential increase to two integral components of 5G networks: 5G small cells and massive multiple-input multiple-output (MIMO) antennas. 5G small-cell deployments are expected to soon surpass those of 4G, with an estimated total installed base of 13.1 million 5G or multimode small cells by 2025, constituting over one-third of the total small cells in operation. While the energy consumption of an individual small cell is lower than that of a traditional cell, the need for a greater number of small cells to cover a specific area leads to an overall higher energy consumption for the network. Similarly, the adoption of massive MIMO technology involves deploying arrays with numerous antennas at each base station, further contributing to the increased overall energy consumption of 5G base stations in comparison to their 4G counterparts.
- IoT: While individual IoT devices exhibit low energy consumption, the sheer number of connected devices expected in the near future has raised apprehensions about the energy requirements of IoT. A more significant challenge, as experts argue, lies in effectively managing the electronic waste (e-waste) generated by IoT devices. According to a United Nations report, global e-waste is projected to increase to 52.2 million metric tonnes by 2025, with IoT being a key contributor. This is primarily attributed to the widespread use of batteries and semiconductors in connected devices. There are also growing concerns about the unnecessary applications of IoT technology, such as embedding sensors and Bluetooth connectivity in basketballs to track user activity.
- AI: While AI holds promise for reducing the long-term energy needs of businesses, the training process for AI-driven machines currently involves significant energy intensity and time consumption. According to a recent study, the environmental impact of training a single AI-powered vehicle is five times greater than the lifetime emissions of an average car. AI-driven machines undergo training using deep learning, a process that entails handling massive datasets, leading to a substantial rise in computing and power demands. For example, in the case of an AI-powered chatbot like ChatGPT, billions of written articles are processed to enable the system to grasp the meaning of words and comprehend sentence structures.
Sustainable energy management strategies
- Powering tower sites through renewable energy solutions: With the recent decline in solar panel prices, solar energy has emerged as a financially viable solution to meet the energy needs of telecom towers in rural areas. Meanwhile, in urban settings, rooftop-mounted solar panels can effectively power telecom infrastructure. Additionally, biomass can be used at rural sites experiencing average loads exceeding 5 kW and facing grid outages of over eight hours. In some cases, the use of hydrogen fuel cells and wind energy may also be considered to meet electricity requirements.
- Deploying-energy efficient cooling equipment in data centres: Given that around 40 per cent of the total energy consumed by a data centre goes towards maintaining a temperature-controlled environment, installing energy-efficient cooling equipment can help significantly reduce power consumption. The use of advanced chiller equipment with low-global warming potential refrigerants can reduce the annual electricity consumption of a data centre by up to 35 per cent. Further, smart cooling solutions powered by predictive analytics can help data centre managers optimally allocate power for different periods. The energy requirements of data centres can also be brought down by replacing physical servers with virtualisation technologies such as software-defined networking, network function virtualisation and cloud computing. Further, data centre operators should also reduce the dead server space prevalent in their facilities. According to industry estimates, nearly one-fifth of all servers in a traditional data centre remain vacant and unused.
- Leveraging virtualisation to optimise cloud computing energy requirements: To reduce the energy quotient of cloud computing, companies can deploy virtualisation technologies to run multiple virtual machines on a single physical server, reducing the number of physical servers and the associated energy consumption. In addition, virtualisation technologies can be used to automate the scaling of computing resources, allowing organisations to adjust their capacity as needed to meet changing demands. Enterprises can also adopt cloud services that offer carbon offsetting programmes, allowing them to offset the carbon emissions associated with their use of cloud technology by supporting projects aimed at reducing carbon emissions.
The way forward
The Indian telecom industry has proactively implemented several measures to reduce the energy footprint of an ever-expanding digital infrastructure. These include installing diesel-free sites, powering network equipment through renewable energy sources, adopting innovative cooling methods such as fan coil unit cooling, DC air-conditioning, rack cooling, free cooling units and natural cooling units, replacing indoor base transceiver stations (BTSs) with outdoor BTSs and promoting tower sharing, leading to significant fuel cost savings in network operations.
However, the significant upfront investment needed to deploy these energy solutions presents a formidable challenge in reducing the telecom sector’s energy requirements, particularly when companies struggle to cover operational costs. Many renewable energy sources require additional storage technologies, further exacerbating the overall capital expenditure burden. Therefore, in addition to the industry’s efforts to reduce its energy footprint, the government should consider ways to provide financial incentives to telecom infrastructure companies for energy reduction. This could involve implementing measures such as tax holidays and accelerated depreciation to offset the additional capex associated with the implementation of renewable energy solutions.
