Ways for Quantum Computing to help fight climate change

The biggest battle we will be, despite ourselves, forced to win in the forthcoming years will be against climate change. World Health Organisation (WHO) estimates that between 2030 and 2050 climate change will cause approximately 250,000 additional deaths per year [1] and if we take a look at what is already happening we cannot be surprised at all: in January 2021, the National Center for Environmental Information (NCEI) published a Global Climate Report [2] regrouping the significant climate anomalies and events of that single month (see Fig. 1). The world is having a collection of new records in terms of global land and ocean temperature. The situation is dramatic and is getting even worse this year.

Fig. 1 — NOAA — Global Climate Report — January 2021

While it is more or less clear to everyone what we could do in order to limit individually our impact on the environment —prefer the second-hand market, sort our waste, use public transportation whenever possible, decrease our consumption of meat and animal by-products, avoid single-use items [3]—, the way to exploit our most powerful weapon, science, in order to create a large-scale and collective change is not yet explored enough. Among the several branches of science, one which seems particularly promising is quantum technologies. The scope of this short article is to summarise the potentially impactful ones for tackling climate change, trying also to give an idea of the timescales for their realization and their contribution to real-world problems.

Chemistry
The first promising application in the near term for quantum computing is in chemistry, more specifically, in the simulation and the discovery of molecules having peculiar and useful behaviors. Quantum-computing tools could finally solve the problem of many-body quantum-mechanics simulations, having the potential to fundamentally change computational chemistry and material science. This is due to the quantum nature of molecules that can find in quantum simulation in a faster and more precise way to be described [4]. Simulating large complex molecules is a hard computer-intensive task. To give an idea of the timescales of these kinds of problems, if we simulate a molecule with 10 atoms and we suppose that this calculation takes a minute, one with 11 would take two minutes, one with 12 atoms would last four minutes, and so on. It is crystal clear from this pair of examples that, unfortunately, the time to perform those simulations scales exponentially with the number of atoms: simulating a molecule with 70 atoms would take more than 13 billion years, which is the lifetime of the universe. One can think that simulating the behavior of such a complex system could not be of any practical use, but the reality is even more complex. For instance, even solving exactly the dynamics of a common protein such as serum albumin, C₁₂₃H₁₉₃N₃₅O₃₇, would involve the model of a molecule composed of 378 atoms [5].
The applications of chemistry simulations on problems of everyday life would be astonishing, resulting not only in a revolution from a medical point of view but also in a significant reduction of greenhouse emissions. Some of the most promising topics falling in the latter category are discussed below.

Fertilizers. Nowadays, in order to feed the growing population (8 billion) and the consequentially growing farm animals population (more than 70 billion [6]), we inevitably need to produce a huge amount of fertilizer. This product is currently done by means of a high energy-consuming method called the Haber-Bosch process, developed in the early 1900s by Fritz Haber and later modified to become an industrial process by Carl Bosch. Since then, this process did not change significantly, and it takes so much energy to be considered responsible for around 2% of the global CO₂ emissions [7]. It is then easy to understand how great the discovery of a new production process would be for humankind, both for climate and economy. Some signs of progress involving quantum computing in order to replace the Haber-Bosch process in the production chain have been done in the last years [8], but these algorithms still require way too many qubits and around 10 million quantum operations, so they are clearly not prone to work on currently available NISQ devices and then they still represent a long-term application.

Carbon Capture. One of the most important catalytic processes to crack in order to reduce the greenhouse gases in the atmosphere is the binding and transformation of carbon dioxide. As a matter of fact, its infrared absorption properties make it a major factor in climate change [9]. Nonetheless, it should be, in principle, captured naturally by oceans and trees, but our production of this gas has exceeded these natural capture rates by several decades.
The existing catalytic processes to capture CO₂ are currently based on expensive precious metals and so we would need a cheap and readily available substitute. We have an almost infinite plethora of possible molecules that could fulfill this task, but testing each one of them is almost impossible without the possibility to efficiently and fastly simulate the properties of these candidates. The role of quantum computers could be crucial for these simulations in the medium term and investments in this direction already exist: for instance, in 2020 the fuel company TotalEnergies announced a partnership with Cambridge Quantum Computing with the aim of improving materials for CO₂ capture [10, 11].

Energy Consumption
In the last years, the idea that a quantum advantage should be searched only in a speed-up or in the resolution of problems until now impossible to treat classically has spread, but a very important advantage has been so far almost ignored: quantum computers could possibly provide a significant reduction in terms of the total energy needed to perform a given calculation, and such a lower requirement would translate in a drastic reduction of the footprint of computing. Just to give an idea of what we mean, we try in the following to compare the performances of a quantum and a classical computer. The energy consumption for these devices is estimated through the following expression [12]:

Energy consumption = Operating time × Power use.

In order to compare the performances of a quantum and a classical computer, we base estimations on the (controversial) Google’s quantum supremacy experiment [13]. In this article, Google claims that their superconductive quantum Sycamore processor took about 200 seconds to sample one instance of a quantum circuit a million times, while this task would have needed at least 2.5 days to be completed on the classical Summit supercomputer [14]. Assuming the good approximation of 25 kW of power usage for the quantum computer [12], including the cooling process and all operation electronics, and 13 MW of power usage for the classical counterpart [15], we obtain:

Quantum: 0.056 h × 25 kW = 14 kWh,
Classical: 60 h × 13,000 kW = 780,000 kWh.

In this specific problem, the quantum computation costs 557,000 times less energy than respect to the classical one.
It is worth noticing that even if the calculation would have lasted the same time, the quantum computer would have, anyway, consumed 520 times less energy. Of course, we still need to wait years to have quantum computers that could be compared to their classical counterparts taking into account the resolutions of different problems, so a significant reduction in the emission-related computation is still years far. Nonetheless, the orders-of-magnitude difference between the energy needed by the two computers makes this a key topic for the climate.

Climate modeling
Climate models are usually based on systems of Partial Differential Equations (PDEs) in which the equations represent the mathematical formulation of the physical laws governing oceans and the atmosphere. Because of the complexity of the system rising from the interactions between a high number of phenomena, these mathematical models can only be solved under strongly simplifying assumptions, which are a priori decisions about which physical processes are important for the precise scope [16]. Nowadays, the most complex and precise description of the climate system is given by General Circulation Models (GCMs), which basically represent the first successful attempt at climate modeling [17], and in order to obtain the solutions of these PDEs, equations developed they have to be transformed into numerical models that can be handled by a classical computer. Unfortunately, climate modeling is currently necessary to have an understanding of how climate and quantum computing could help us, solve PDEs, to obtain more precise and complete solutions in a reasonable time. These calculation could allow an understanding of how climate has changed in the past, hoping to identify any underlying trends to deal with them, in order to unveil any possible solutions as soon as possible.

Fuel-saving
Even before the spreading of computers and GPS, some gimmicks to save time and fuel have been devised: it is a well-known fact that UPS drivers almost never turn left [18]. Left turns are indeed less safe and time-waster in a right-based driving system such as the one adopted in a major part of the world. This weird trick due to drivers' insight avoids emissions equivalent to over 20,000 passenger cars. At the beginning of 2008, UPS started its investments in a more scientific route optimization algorithm, the so-called On-Road Integrated Optimisation, and Navigation (ORION). Without surprise, in 2012 ORION confirmed what the drives guessed and applied since the 1970 and pushed further UPS fuel-saving reaching over 38 million liters per year [19]. The kind of optimization problems ORION deals with, known as Traveling Salesman Problem (TSP) and Vehicle Route Problem (VRP) (see Fig. 2), increases faster than exponentially: for only 16 delivery stops already over 20 trillion possible routes can connect them all.

Fig. 2 —Schematic illustration of TSP and VRP [20].

Until now, ORION is not able to solve the TSP but it bases the optimization process on Machine Learning (ML) using training sets based on years of data to identify route combinations that are efficient enough [21]. Anyway, dealing with a brand new set of routes makes ML useless and imposes solving such a problem from scratch. Once again, so many possible trajectories should be tested that such a problem would be unsolvable even for the most powerful supercomputers. Quantum computing might once again represent the key to facing this need, as supply chains and logistics become more fundamental for our society. Nonetheless, the work on this topic is still in the early stages, so that is hard to say if quantum would be really more efficient than classic in this optimization task. Anyway, some interesting results have already been obtained using a quantum-classical hybrid approach, based on quantum annealing technology [22, 23].
Many other ways to optimize the energy consumption of transportation means could be discovered as soon as the resolution of PDE with a quantum computer will be cracked. New airplane designs saving each year tons of CO₂ would be then easy to find, leading us towards the green revolution we already desperately need.

At ColibrITD

At ColibrITD, we strongly believe that each one should play their part to fight climate change. Understanding how to exploit quantum technologies and help their development to this aim is one of our main goals.

References:
[1] Climate change and health (who.int)
[2] January 2021 Global Climate Report | National Centers for Environmental Information (NCEI) (noaa.gov)
[3] How to Reduce Your Carbon Footprint — A Year of Living Better Guides — The New York Times (nytimes.com)
[4] A quantum-computing advantage for chemistry — PubMed (nih.gov)
[5] Serum albumin (1–24) | C123H193N35O37 — PubChem (nih.gov)
[6] Opinion | We Will Look Back on This Age of Cruelty to Animals in Horror — The New York Times (nytimes.com)
[7] Current and future role of Haber–Bosch ammonia in a carbon-free energy landscape — Energy & Environmental Science (RSC Publishing)
[8] Accuracy and Resource Estimations for Quantum Chemistry on a Near-term Quantum Computer
[9] Quantum computing enhanced computational catalysis (aps.org)
[10] Quantum computing for innovative climate change solutions | World Economic Forum (weforum.org)
[11] Total is exploring quantum algorithms to improve CO₂ capture | TotalEnergies.com
[12] Quantum technologies for climate change: Preliminary assessment (arxiv.org)
[13] Quantum supremacy using a programmable superconducting processor | Nature
[14] On “Quantum Supremacy” | IBM Research Blog
[15] US Dethrones China With IBM Summit Supercomputer | Tom’s Hardware (tomshardware.com)
[16] Introduction to climate dynamics and climate modelling
[17] The general circulation of the atmosphere: A numerical experiment — Phillips — 1956 — Quarterly Journal of the Royal Meteorological Society
[18] Why UPS trucks (almost) never turn left — CNN
[19] Analytics Success Story: UPS’s ORION Project | Introduction to Business Analytics and Decision-Making | InformIT
[20] Minimizing the Carbon Footprint for the Time-Dependent Heterogeneous-Fleet Vehicle Routing Problem with Alternative Paths (mdpi.com)
[21] How Quantum Computers Could Cut Millions Of Miles From Supply Chains And Transform Logistics (forbes.com)
[22] A Hybrid Solution Method for the Capacitated Vehicle Routing Problem Using a Quantum Annealer (arxiv.org)
[23] University of Warsaw: Solving Vehicle Routing Problem Using Quantum Annealing — YouTube