Low extension rates in continental rifts lead to very long recurrence times of large earthquakes, which impede detailed seismic hazard analysis. It is, however, clear that rift earthquakes can be devastating, as, for instance, in the East African Rift System Fig. b;. The MS 7.4 Kasanga earthquake, one of the largest ever recorded in Africa, rocked the Rukwa Rift area in 1910 and caused submarine slumps that broke telegraph cables off the coast of Mozambique (see also Sect. 3.3.2). Casualties were limited, though, most likely due to the style of local buildings at the time . Other notable earthquakes with M>6 in the last century include the 1928 Subukia earthquake in the Kenya Rift (MS=6.9); the 1966 Toro earthquake in Uganda (MS=6.1); and the 1960 Hawassa earthquake (MS=6.1) and the 1989 Dobi graben event (MS=6.5), both in Ethiopia . The Afar Triple Junction region is particularly earthquake-prone .

Significant continental rift-related seismicity is also found outside Africa. In western Europe, the year 1356 saw the total destruction of medieval Basel, near the southern tip of the Upper Rhine Graben (; Fig. c). Smaller earthquakes also occur regularly along the Lower Rhine Graben in Germany and the Benelux e.g.. The Aegean back-arc system is prone to rift-related quakes as well, from the Corinth rift and the area further east around Athens to the western coast of Turkey e.g.. Other examples of active continental rift systems posing hazards are the Gulf of Suez rift and the Trans-Mexican Volcanic Belt, the latter of which contains various major cities such as Mexico City and references therein. The Siberian Baikal Rift system is also seismically active, with many earthquakes exceeding magnitudes of 6. This is due to the fact that the rift is situated in a cold lithosphere, where the brittle–ductile transition is deep and fault planes have a larger extent. The largest recorded Baikal Rift earthquake even reached a magnitude of MW=7.8 the 1957 Muisk event;, However, as population densities are low in the area, risks are moderate . Conversely, in China, exceptionally strong and deadly rift-related earthquakes are known to occur in the Shanxi Rift, southwest of Beijing . This rift, which has been active since the Pliocene, was the site of the 1303 Hongdong and the 1556 Huaxian earthquakes, with magnitudes of MW 8 or higher, which caused over 470 000 and 830 000 fatalities, respectively .

Numerous earthquakes are recorded at mid-oceanic ridges (Fig. ), yet these pose only limited hazards due to their distance from population centres. There are situations where oceanic ridges are close to continents, or enter into continental lithosphere, and where significant earthquakes occur, such as in the Gulf of California or the northern tip of the Red Sea . A special case is found in southern Iceland, where the emerging Mid-Atlantic Ridge runs close to the area around the capital of Reykjavik and where historic earthquakes, such as the MS=6.0 event in 1706, have caused death and destruction e.g..

The above examples concern actively deforming rift systems. However, rift-related intra-plate earthquakes can also occur without active plate tectonic motion, i.e. in stable continental regions away from active plate boundaries that have not seen significant deformation over the recent geological past . A notorious example is the New Madrid Seismic Zone in the central USA, which represents a concentration of seismicity linked to the tectonically inactive Palaeozoic Reelfoot Rift and was the locus of the devastating 1811–1812 earthquakes with magnitudes over MW 7 . Other examples include the 1663 Charlevoix earthquake along the St. Lawrence rift zone in Quebec and the 1845 and 1888 earthquakes in the Río de la Plata region, Argentina, which were linked to the inactive Mesozoic Quilmes Trough . Given that large swathes of the Earth's continental lithosphere are under some level of stress World Stress Map;, and that inactive rift systems may act as inherited weakness zones and are widely distributed Fig. , large earthquakes may occur at unexpected locations e.g..

Furthermore, intra-plate earthquakes can strike along rifted margins, which are often somewhat misleadingly referred to as “passive margins” (see Sect. 2.1). Ongoing deformation caused by various processes, such as magmatic underplating, plate convergence, or glacial rebound, can trigger the reactivation of former rift faults to generate significant earthquake activity. A notorious example is the magnitude 7.2 Grand Banks earthquake offshore Newfoundland in 1927 , which caused a large submarine landslide and an associated tsunami (see Sect. 3.3.2).

Research and monitoring allow us to assess hazards, vulnerability, and risk in specific rift systems. Seismic surveys and geophysical analyses, satellite-imagery-based InSAR studies, and measurements of the state of stress in boreholes provide insights into the local geological setting and the dynamic deformation causing earthquakes . Important constraints on earthquake recurrence time can be deduced from historical literature e.g., from dating of known earthquake-prone faults by investigating sediment deposits adjacent to faults and palaeoseismic trenching analysis and references therein, from directly dating the fault surface with cosmogenic nuclide analysis e.g., or by archeological research . Relevant data are collected and made available by various organisations, such as the Malawi Seismogenic Source Model in the East African Rift System , the EU-funded EPOS Seismology Consortium, and the European Facilities for Earthquake Hazard and Risk (EFEHR) Consortium, or on a global scale by the Global Earthquake Model (GEM) project. These datasets are crucial for disaster management planning and for the safe development of specific industries, especially those involving nuclear power production and fluid injection (geothermal energy projects and unconventional hydrocarbon production). For instance, the geothermal fluid injection tests in Basel, at the southern tip of the Upper Rhine Graben, caused slip along pre-stressed faults and minor earthquake damage in the city centre . Moreover, extensive unconventional hydrocarbon production from sedimentary deposits in a variety of inactive (rift) basins in the USA causes regular seismicity with magnitudes over 5.5 MW e.g..

Volcanic hazards generate the most risk in rift systems prior to the break-up stage, when the system evolves mostly in subaerial conditions, and where the fertile volcanic soils provide excellent farmland to sustain large human populations. The most prominent examples of such rift-related volcanism may be found along the highly magmatic eastern branch of the East African Rift, where, among others, Mount Kilimanjaro is situated (Fig. ). In this context, the Afar Rift at the northernmost end of the eastern branch is a special case, as some authors characterise it as an embryonic (subaerial) mid-oceanic ridge . This highly volcanically active region is mostly a desert at present, with parts situated down to 160 m below sea level , but contains considerable populations that live with the risks of frequent volcanic eruptions. For instance, although the 2008 Alu-Dalafilla eruption did not cause any human or economic losses due to the remote location of the volcano , the 2011 eruption of the Nabro volcano caused several fatalities . Moreover, the 2011 Nabro eruption resulted in the expulsion of 1.5 Mt of SO2, ranking it as the largest eruption since Mount Pinatubo's in 1991 , and disrupted air travel in the region . The western branch of the East African Rift System also has a moderate amount of magmatism, specifically along the linkage zones between individual rift segments that allow easier fluid migration (e.g. ; Fig. ). As large populations are situated along the shores of the various big lakes of the western branch of the East African Rift System, the risk posed by volcanism is significant e.g.. A recent example is the 2002 devastation of the city of Goma in the Democratic Republic of Congo by a volcanic eruption of the nearby Mount Nyiragongo .

The Aegean back-arc system is another location that is prone to continental volcanism, with the 1600 BCE Thera eruption on the modern-day Greek island of Santorini being perhaps the most notorious volcanic event in the recent geological past of the Mediterranean. This especially violent eruption, one of the largest in recent human history, devastated the island and sent tsunamis throughout the eastern Mediterranean , and is thought to have been instrumental in the collapse of the previously flourishing Minoan civilisation on nearby Crete and references therein. The Trans-Mexican Volcanic Belt and its numerous volcanic eruptions also pose serious hazards to the various major cities in the area, including Mexico City and references therein. The Taupō Rift Zone, where the Havre Trough back-arc rift system enters the continental lithosphere of New Zealand, is another site of ongoing volcanism that witnessed a super-eruption as recently as 26.5 ka . Other volcanic zones in continental rift settings are the Eifel, Massif Central, and Eger Graben areas of the European Cenozoic Rift System, which are less known due to the absence of present-day eruptive activity (Fig. ). However, the last eruption in the Massif Central occurred only around 8 ka , whereas large-scale volcanism in the Eifel area at the northern tip of the Upper Rhine Graben lasted until some 10 ka . Also, the Eger Graben further to the east exhibits geologically recent volcanic activity . A caveat is that this rift-related volcanism in western Europe was possibly caused by mantle activity induced by the Alpine orogeny, the tectonic event that is thought to have induced the rifting in western Europe in the first place . Even so, there is potential for future volcanism in these areas .

After the establishment of continental break-up, oceanic rift systems tend to be submerged, which reduces volcanic hazards and their impact. Only when a submerged spreading axis is close to the continent, as in the case of the Gulf of California, could there be some risk of an eruption affecting human populations. However, in specific locations, spreading ridges (or spreading-ridge-related volcanoes) are subareal and inhabited. Examples are the volcanic Azores islands on the triple junction between the North American, European, and African plates and, most famously, Iceland, situated on the Mid-Atlantic Ridge and site of the 2010 Eyjafjallajökull eruption that caused massive disruptions to global air travel . Next to typical hazards experienced by inhabitants of volcanic islands such as the Azores, the population of Iceland also has to contend with eruptive melting of glaciers sitting on top of their island's volcanoes causing truly massive flash floods that devastate anything in their path so-called jökulhlaups;. Interestingly, the growth and decline of glaciers on Iceland is linked to the intensity of volcanic activity, as the associated increase and decrease in glacial load reduces or enhances decompression melting . A similar correlation between glacial extent and volcanism has also been noted in western Europe .

In addition to the direct volcanism-related hazards such as explosive eruptions, magma-induced earthquakes, lava flows, pyroclastic flows, and lahars/mudflows that can lay waste to large swatches of land, magmatic activity in rift settings is accompanied by substantial hydrothermal circulation and degassing e.g.. These can lead to the development of mud volcanoes and geothermal fields that can pose serious hazards to local populations e.g.. Hydrothermal flow can alter the state of stress in the subsurface, causing the activation of faults, and degassing can pose major local threats as well, in particular when it comes to CO2; i.e. being heavier than air and scentless, CO2 can accumulate in depressions and suffocate unaware people and livestock . Furthermore, CO2 (and other gases) can accumulate in vast volumes in the deeper parts of the water column of lakes along rift systems. These dissolved gases in such meromictic lakes may violently escape when the delicate equilibrium in the water column is disturbed due to an earthquake, a landslide, or some (minor) eruptive activity, or possibly even without external trigger at all . The CO2 released during such limnic eruptions may then spread like an invisible wave through the surroundings, suffocating anyone in its path . A tragic example occurred at Lake Nyos, one of a series of volcanic lakes situated along the Central African Shear Zone in Cameroon, where over 1700 people perished in 1986 . Similar dangers may be lurking in the great lakes of the East African Rift System, especially as these lakes are much larger than those in Cameroon. A lake of particular concern is Lake Kivu, with millions of people living in its vicinity e.g..

Magmatism in rifts poses severe risks, but these can be mitigated to some degree. Induced degassing of meromictic lakes such as Lake Nyos and Lake Kivu reduces the risk of limnic eruptions and even allows the production of CH4 used for local energy needs ; see Sect. 4.2. While exploring for resources in such lakes, it is crucial to monitor the equilibrium in the water column so as not to accidentally cause a limnic eruption. Furthermore, there are seismic networks actively monitoring earthquakes that could kick off a limnic eruption in meromictic lakes . We thus have some means to actively mitigate the risk of limnic eruptions, but preventing volcanic eruptions is an unfeasible proposition, so at present we can only prepare and react when they happen. Similarly to meromictic lakes, monitoring of seismic activity, variations in degassing, or changes in ground elevation that may hint at disturbances in the magma chamber and the possibility of an imminent volcanic eruption are key monitoring methods applied in various volcanic rift settings . A big challenge may be the recurrence time of eruptions in rifts that may run in the hundreds of years e.g.. However, in some cases, it is possible to mitigate the potential destruction caused by lava flows by diverting their paths, as was done during the 2023 and 2024 Grindavík eruptions that threatened the important Svartsengi geothermal plant and the famous Blue Lagoon spa . Another intriguing example occurred on the Vestmannaeyjar archipelago, just south of Iceland, where a lava flow on the main island of Heimaey in 1973 threatened to block the harbour entrance. Taking into account that fishing was the economic lifeline of the inhabitants, the lava flow was successfully prevented from advancing too far by spraying its front with large volumes of seawater .

Rifting causes the development of topography and thus of unstable slopes, either subareal or submarine, that can collapse in mass-wasting events. In subareal situations, footwall uplift of major normal faults and subsequent erosion can form impressive escarpments, such as those along the East African Rift System and the Red Sea–Gulf of Aden system, which in some cases represent some 2 km of topographic difference between rift basins and their shoulders e.g.. Steep slopes and escarpments can be destabilised by mechanical weakening or by ground-shaking leading to frequent landslides . The most voluminous landslides in the East African Rift are caused by ground-shaking due to earthquakes and by precipitation-induced weakening . Particular areas at risk are the escarpments bordering the Ethiopian Highlands and large parts of the western branch of the East African Rift , for instance, along the shores of Lake Kivu .

Other, often more impactful mass-wasting events occur offshore, at the continental slopes of rifted margins. Here, large volumes of sediments accumulate over time, since the world's largest rivers drain towards rifted margins. Importantly, rifted margins in high latitudes receive major sediment input as a result of glacial erosion during ice ages. These large volumes of sediment are deposited at the angle of repose so that a slight destabilisation, either due to additional sedimentary loading, earthquakes, or sediment weakening, can cause a catastrophic collapse and massive submarine landslides. Various instances of such collapses and consequent landslides have been recorded by the severing of communication cables over the seafloor e.g.. Famously, the magnitude 7.2 Grand Banks earthquake offshore Newfoundland in 1927 caused a massive submarine landslide, including turbidite flows that travelled over a distance of over 1000 km at speeds ranging between 60–100 kmh-1, thereby severing telegraph cables . Communication cables can be replaced, but tsunamis are much more impactful hazards generated by submarine slides, such as the one that resulted from the 1927 Grand Banks earthquake, which claimed 28 victims . However, the most well-known submarine landslide is the Storegga landslide that occurred along the Norwegian rifted margin in ca. 8000 BCE . The resulting tsunami drowned the coasts of Norway, the Shetland and Orkney Islands, eastern Greenland, and northwestern Europe down to the southern North Sea . Many other submarine landslides and associated tsunamis are known from the northeastern Atlantic e.g. and from rifted margins around the globe (e.g. ; Fig. ). Since many large cities are located along the coast near rifted margins, the impact of tsunamis induced by large submarine slides such as those offshore Norway would be catastrophic. We should, however, emphasise that the recurrence intervals for these large submarine landslides are very broad . Still, climate change and associated potential destabilisation of rifted margins increase the risk of landslide tsunamis in the North Atlantic and along other rifted margin settings.

Mineral deposits are anomalous concentrations of minerals in rocks or sediments, while ore deposits are those deposits that are economically viable to extract . The mineral deposits related to rift systems can be divided into various categories depending on when and how they formed (e.g. ; Fig. ).

Pre-rift mineral deposits are of great importance in rift systems. They come in many shapes and forms, depending on the pre-rift (Stage 0) geological history of the specific setting. The rocks containing pre-rift deposits are exhumed by the combined action of extensional tectonics and erosion at the rift shoulders, along individual rift blocks, or along (uplifted) rifted margins. Much of the mineral wealth of the countries along the East African Rift System is based on such deposits exhumed during rifting e.g.. Other examples of pre-rift deposits are those which are (historically) exploited along the Cenozoic European Rift System (e.g. the Vosges and Black Forest mountains and the Massif Central), where the Hercynian metamorphic basement is exposed e.g..

Sedimentary mineral deposits are generated by sedimentary processes. One type of sedimentary deposit, called “(palaeo)placers”, is found in clastic sedimentary environments. Placer deposits form through the breakdown of bedrock and the sorting and concentration of relatively heavy minerals, such as diamonds, gold, and platinum-group minerals, by various sedimentary processes mostly fluvial;. Deposits thus form at shorelines and in rivers and are found, for example, in the East African Rift System and along the rifted margins of eastern Africa and the Russian Arctic . High-grade diamond deposits are found and exploited offshore Namibia . Not all placer deposits are economical, but they can still provide insights into the location of their (pre-rift) origin sites through provenance analysis .

Hydrogenetic deposits, which are chemically precipitated from surface waters, can also be found in rift environments, for example, ancient or actively forming evaporites. Evaporites develop in compartmentalised basins through solar evaporation and can be mined to produce salt for human consumption or industrial use, as is the practice in present-day Afar and in Lake Magadi in Kenya . Depending on the origin of the salt and fluids marine or hydrothermal, or even from serpentinisation of exhuming mantle material, referred to as “dehydrates”;, there may also be important concentrations of potash, lithium, trona, and other minerals to be extracted . For instance, phosphorites are phosphorus-rich hydrogenetic deposits that (bio)chemically form in the shallow marine environment along rifted margins i.e. on the continental shelf;. Given the projected scarcity of phosphorus, a component of fertiliser , these relict and recent offshore deposits will be of great interest to the global community in the near future.

Sediment-hosted hydrothermal ore deposits form in sedimentary rift basins both during and after rifting. Hydrothermal fluids circulating through the basins leach metals from deeper in the basin, transport them upwards, and deposit them in sulfide minerals when reaching conditions that trigger metal release . Temperature, salinity, pH, and redox state all affect the solubility of metals in these fluids e.g., making heated saline waters an efficient mineralising fluid and colder, reduced rocks a possible host .

Sediment-hosted mineral deposits are the largest global resource of lead and zinc , which are both base metals with great importance to the global energy transition . The majority is found in clastic-dominated (CD)-type deposits formerly known as SEDEX deposits; that formed during deposition or diagenesis of the marine siliciclastic host rock through subseafloor replacement of barite with lead–zinc sulfides syngenetic to diagenetic mineralisation;. These mostly Proterozoic and Paleozoic mineralisations occur close to syn-sedimentary faults e.g., which could have provided permeable pathways focusing upward fluid flow e.g.. Large deposits have been found in the Carpentaria Zinc Belt in Australia, the North American Cordillera, Russia, and China e.g.. The precursory sediment-hosted barite deposits are also mined in China, India, and the USA e.g..

Contrary to CD-type deposits, Mississippi Valley-type (MVT) lead–zinc deposits formed much later than their host rock (up to tens of millions of years; epigenetic mineralisation). The mostly Phanerozoic MVT deposits are found in broad rift basins with thick sequences of shallow marine sediments including platform carbonates that host the mineralisation . These basins have generally been affected by inversion e.g., with high topography and/or tectonic loading possibly having driven long-distance fluid flow . By finding correlations between geophysical and geological observational datasets and the locations of known CD-type and MVT deposits, researchers try to limit the exploration space for these deposits e.g.. For example, found that all giant deposits lie within 200 km of the edges of cratonic lithosphere.

A big supplier of copper is the sediment-hosted stratiform copper (and cobalt) deposits as found in the Kupferschiefer field in the Zechstein Basin (Poland and Germany), the Central African Copperbelt, the Kodar–Udokan basin in Siberia, and the White Pine deposit in the USA. These deposits mostly occur at stratigraphic and redox boundaries between terrestrial syn-rift redbed sandstones (metal source) and overlying shallow-marine or lacustrine organic-rich shales host rock;. Such successions formed in the intracontinental basins of failed rifts, with syn- or post-diagenetic mineralisation by oxidised saline fluids during the sag or even the inversion phase .

Intracontinental basins also contain several types of sediment-hosted uranium deposits e.g.. For example, epigenetic-unconformity-related deposits such as in the Athabasca Basin, Canada; are found at major unconformities between the basement and overlying redbeds . Tabular and roll-front deposits have distinct blanket and U-shaped geometries, respectively, and are formed by lower-temperature oxidised fluid circulating in the sandstones .

An example of magma-related hydrothermal ore deposits, or volcanic-hosted or volcanogenic massive sulfide (VHMS or VMS) deposits, formed and currently form through venting of hydrothermal fluids in and at the seafloor creating, for example, black and white smokers;. The first seafloor hydrothermal activity was discovered in the Red Sea oceanic spreading centre , and many more sites have been identified since. The majority of active VMS deposits occur along mid-oceanic ridges, where mostly seawater circulates in the subsurface, depositing polymetallic sulfide minerals that contain, amongst others, copper, zinc, lead, gold, and silver e.g.. Some fossil VMS systems are found exposed after being obducted by subsequent convergence tectonics. The Oman ophiolite is a well-known example, but the Troodos Ophiolite on Cyprus is world-famous since its VMS deposits have been exploited for thousands of years to the degree that copper (cuprum, in Latin) may have been named after the island . Currently active VMS systems are common along mid-oceanic ridges but are also found in the (incipient) spreading centres of Afar e.g. at Dallol;.

Diamondiferous kimberlites are a well-known type of magmatic ore deposit that contain diamonds. Kimberlites are rocks formed from highly volatile ultramafic eruptions in cratons, shields, and mobile belts , which can transport diamonds that are formed at large depths within the mantle (>150 km) to the surface . Recently, correlated kimberlites with periods of dispersal of the continental plates, with a lag of about 30 Myr between break-up and kimberlite volcanism. Also, the emplacement of the Argyle lamproite diamond deposit in the intracontinental rift in the Halls Creek Orogen (Australia) was probably driven by the break-up of the Nuna supercontinent . Other magmatic ore deposits linked to rifting are layered intrusions, such as those of the Bushveld Complex in the Kaapvaal Craton, which likely formed in a back-arc setting and contain, amongst others, copper, nickel, and platinum-group metal deposits (Fig. ). Moreover, syn-rift carbonatites, found, for example, along the East African Rift System and in the Bayan Obo deposit in China, produce the majority of rare-earth elements and niobium . Carbonatite volcanism is rather rare in the geological record, and the Ol Doinyo Lengai in the Tanzanian part of the East African Rift System is the only currently active volcano that produces carbonatites . Chromite can be found both in layered intrusions like the Bushveld Complex e.g. and in mid-oceanic-ridge and back-arc ophiolites e.g. the Troodos, Oman, and Kempirsai ophiolites;.

Rift environments contain a variety of aggregate mineral resources that can serve as construction materials. For instance, deposits of sand and gravel in rift environments are of great importance for construction projects, since they are used for roads and building foundations and for the preparation of cement and concrete . Clay deposits can be used for brick production, and sedimentary rocks are used for the construction and decoration of buildings. For instance, Triassic Buntsandstein outcropping on the rift shoulders of the Upper Rhine Graben has always been a popular building material in the region . Another application is the use of Kieselkalk, which was deposited on the European rifted margin prior to the Alpine orogeny (https://www.strati.ch, last access: 11 July 2024) as railway track ballast . Volcanic deposits found in rift settings are also of interest to construction. Imported basalt blocks have been applied to reinforce waterworks in the Netherlands , and volcanic rocks were used to construct the old town of Clermont-Ferrand and its iconic black cathedral in the Massif Central . Eruptive materials in the East African Rift System are used for various local construction purposes but have the drawback that their characteristics are rather unpredictable due to their heterogeneous composition and grain size variations . Even so, cinder material is a reasonable substitute for scarce gravel in eastern Africa .

Another highly valuable resource that can be produced in rift systems is helium gas, which is used for a wide range of applications ranging from healthcare to space flight . A highly strategic non-renewable resource, helium is normally extracted from rocks in felsic cratons, which accumulate the gas that is generated as a by-product of uranium and thorium decay . However, unusually high helium fluxes in the Tanzanian part of the East African Rift System have recently caught the attention of exploration geologists, and large helium reservoirs have been discovered . The cause of these high fluxes is the increased heat of the rift system, which releases the helium from cratonic rocks, after which it accumulates in porous reservoir rocks from which it can readily be produced .

Hydrocarbons presently represent the most well-known energy resources produced in rift systems, with massive oil and natural gas deposits found in petroleum provinces such as those in the North Sea rift and UK–Norwegian margin and in the Gulf of Mexico (e.g. ; Fig. ). Rift systems provide ideal environments for the development of conventional petroleum systems. Firstly, they allow the development of hydrocarbon source rocks, typically fine-grained marine or lacustrine sediments rich in organic material, in restricted rift basins with fast subsidence, limited water circulation, and low O2 content, so that the organic matter is preserved and only minor amounts of other sediments are deposited in these sediment-starved environments . Rapid subsidence in rift basins also allows the source rock to be buried to a depth of 2–7 km by subsequent sedimentary infill to reach temperatures of ca. 100–250 °C that allow the generation of oil and natural gas . Furthermore, the large variety of deposits in rift systems ensures that there is a high likelihood that reservoir rocks are available, especially in earlier rift stages and near active normal faults Fig. ;. The upward migration of buoyant hydrocarbons from the source rock into reservoir rocks can occur vertically, but pathways such as high-permeability sedimentary rock layers and normal faults in rift systems, in particular where individual rift basins merge, allow enhanced hydrocarbon migration . The presence of impermeable rock layers, such as clays (or evaporites) that generally develop in the later stages of rifting, act as seals or cap rocks, forcing hydrocarbons to migrate around them . However, when this is not possible, as in the case of a tectonically or stratigraphically formed trap structure, a hydrocarbon field can develop, which can be drilled and exploited . It is key that these elements need to be present at the right time and the right place, so a thorough understanding of a basin's geological history is required to successfully exploit its conventional petroleum system(s) .

Furthermore, relatively recent technological developments have allowed the exploitation of unconventional petroleum systems, which include shale oil and/or shale gas and are much simpler than conventional petroleum systems since the source rock is simultaneously the reservoir rock . Large volumes of hydrocarbons remain trapped in the impermeable fine-grained source rocks themselves, and they can be released and produced by directly drilling into the source rocks followed by hydraulic stimulation or “fracking” . Rift settings provide an ideal environment for the development of source rocks, which can be exploited even when no suitable conventional reservoirs, traps, or seals are present in the system. Such unconventional production of shale oil and gas has been booming in the USA since the late 20th century, with various rift basins producing unconventional oil and gas e.g. the Southern Oklahoma Aulacogen;. Although there seems to be good potential in other rift basins around the globe e.g. the North Sea rifts; the Pannonian and Donetsk basins in Europe; or the Songliao, Fuxin, and Bohai basins in China;, these resources have not yet been developed. A notable exception is the Neuquén Basin in Argentina, which contains the highly prolific Vaca Muerta source rock that was deposited in a back-arc rift setting .

Gas hydrates represent another type of non-conventional hydrocarbon resource that is abundantly found along rifted margins and uniquely in the continental Baikal Rift as well . At great depth, or at low temperatures in more shallow marine environments such as in the Arctic, natural gas (commonly generated by biochemical activity in the seafloor) and water can form an ice-like compound referred to as gas hydrate . Massive deposits of gas hydrates are known to exist along rifted margins, and various pilot projects have been undertaken to explore their potential e.g. at the margin of the Canadian Arctic and the South China Sea;.

Finally, hydrocarbons also come in solid form, i.e. as coal, which stems from plant material accumulated and preserved in swamps that is subsequently buried and increasingly enriched in carbon. During this coalification process, the peat first turns into lignite (brown coal), and via a number of intermediate stages the purest type of coal (anthracite) can be formed . Various coal fields have developed in rift settings, especially where subsidence was not too fast so that peat swamps could develop, such as in the marginal basins on both sides of the North Atlantic . Coal layers generate natural gas that can accumulate in hydrocarbon reservoirs, but this gas can also be directly exploited, similarly to unconventional shale gas source rocks coal-bed methane;. Moreover, the various types of coal have been mined for a long time, fuelled the industrial revolution, and remain an indispensable energy source for many countries around the world to this day .

Hydrogen gas (H2) is a clean source of energy, as its combustion generates nothing but water as a by-product. The problem is, however, that present-day H2 production is costly at best (when using green energy) or highly polluting at worst (when using fossil energy) e.g.. However, various sources of naturally occurring H2 exist, of which the most promising is the serpentinisation of (ultra)mafic rocks (e.g. mantle rocks): by reacting with water, these mantle rocks release natural H2 (Fig. ) e.g.. Large amounts of natural H2 are released during the more advanced stages of rifting (i.e. break-up and drifting, when mantle material is being exhumed and serpentinised) (e.g. ; Fig. ), and it is speculated that such natural H2 could have played a key role in the emergence of life on Earth . Of key importance is that water can reach down to the mantle material, for instance, via large normal faults, but complex and deep-rooted fault patterns in rift transfer and transform zones provide improved opportunities for water circulation, as is also the case for magma and hydrocarbon migration (see Sect. 4.2.1). Similarly to conventional petroleum systems, the generated natural H2 from such “hydrogen systems” will have to migrate from the (mantle) source rock to reservoirs in order to be exploited, or it may even be possible to directly drill into the mantle and hydraulically stimulate the mantle source rock in order to serpentinise it . In the case of a natural H2 reservoir, the highly reactive character of H2 means that it can be lost through various (bio)chemical processes. Therefore, reservoirs will ideally have temperatures between 100–200 °C, at which H2 is relatively inert .

Finally, geothermal energy production is a developing industry in rift basins around the world . The thinning of the lithosphere and the rise of hot mantle material towards the surface create an elevated geothermal gradient, and the resulting higher heat flow can be exploited. Geothermal springs in rift basins such as the Upper Rhine Graben have been in use for bathing and healing purposes since at least Roman times e.g.. Furthermore, warm water from shallow aquifers (low-enthalpy systems) can be used for heating greenhouses and other buildings , whereas water from deep aquifers is hot enough for power production high-enthalpy systems;. The drilling targets for geothermal power production pose a challenge, as they may be situated in the deepest part of the rift basins, in contrast to drilling targets for hydrocarbon production that tend to occur higher up in the basin stratigraphy . However, geothermal energy production also provides additional mining opportunities by extracting dissolved elements and minerals from the geothermal fluids e.g. lithium, rare earths, salts;. Various geothermal projects are currently underway in continental rifts e.g. in Soultz-sous-Forêts, France, and Olkaria, Kenya;, with the East African Rift as an obvious target . However, subareal oceanic spreading ridges in Iceland and Afar provide perhaps the greatest potential. In the Icelandic case, the country has such a massive surplus of geothermal energy that energy-intensive industries such as aluminium production have established themselves . The potential in Afar remains untapped as of now, but its geothermal setting, which is similar to that of Iceland, could provide the surrounding regions with copious amounts of green energy in the future .

The best-known temporary geological storage targets are old hydrocarbon fields and porous rock layers, which are plentiful in rift systems. For instance, in western Europe, excellent reservoirs provide temporary storage for natural gas imported from other regions of the world , smoothing supply to meet daily to seasonal changes in production versus demand and providing grid-scale buffering for renewable energy from wind an solar sources, thus improving energy security. Porous rock layers also provide opportunities for Aquifer thermal energy storage (ATES) and high-temperature ATES (HT-ATES). Through ATES, heat can be transferred to water and injected into relatively shallow reservoir layers, where the heat can be effectively stored due to decreased heat loss due to advection (water flow) and diffusion . Of key importance for storage of natural gas and heat in subsurface reservoirs is their widespread availability and excellent permeability, which allows the gas or hot water to be easily injected and extracted .

Another storage target in rift systems involves salt caverns, created by dissolution mining of salt in salt diapirs or by physical mining of evaporite layers (; Fig. ). In contrast to porous rocks, which provide storage space between (sedimentary) grains, caverns and mine shafts are wide, open spaces in the subsurface, which allow rapid injection and extraction of liquid or gaseous resources. Salt is highly impermeable and self-sealing, and caverns provide ideal storage opportunities for strategic petroleum reserves, H2 gas, helium gas, and compressed air . Such storage sites have been developed in the rifted margin of the USA Gulf of Mexico, which contains a large salt tectonic system e.g.. In particular, three commercial compressed-air energy storage (CAES) facilities currently exist in Germany, the USA, and Canada, each exploiting salt caverns . A limiting factor for the use of salt caverns for CAES is the distribution and accessibility of suitable salt deposits, although recent work has shown that compressed air may also be stored in porous media with capacities suitable for national grid-scale energy storage and efficiencies of up to 0.67 . However, care is required to inject compressed air into depleted oil and gas reservoirs due to the potential for a combustible environment at the surface or in the subsurface .

In our push to establish a sustainable economy, we should not only apply the resources available to us but also responsibly process waste products. Where some of these waste products can be recycled, some cannot and need to be permanently and safely stored, which can be done by making use of geological formations found in rift settings.

The most well-known waste product generated by industrial activity is CO2, which is generally released into the atmosphere, where it subsequently is one of the principal causes of global warming. As such, it is essential to reduce the atmospheric CO2 content by capturing and permanently storing it, a process commonly referred to as carbon capture and sequestration (CCS) . For instance, CO2 can be stored in saline aquifers or depleted hydrocarbon reservoirs in rift environments , where it is simply pumped into the reservoir to take the place of the previous hydrocarbons . Buoyant CO2 is initially trapped in the reservoir by an overlying cap rock, but with time CO2 can become more securely stored by other mechanisms such as residual trapping, dissolution, and mineralisation . The possibilities of applying such CO2 storage in depleted hydrocarbon fields in, for instance, the North Sea rift and along the rifted margins of the Gulf of Mexico, China, and Australia are being studied , with a successful CCS project being under way in the Danish North Sea . However, the storage of CO2 in depleted hydrocarbon fields comes with a number of potential challenges, including leakage of the buoyant CO2 due to abandoned infrastructure (e.g. boreholes) and human-induced seismicity during injection/extraction e.g..

In some active hydrocarbon fields, CO2 is also being pumped into the reservoir to increase its internal pressure, thus enhancing hydrocarbon production while storing CO2 , a process referred to as enhanced oil recovery (EOR) or carbon capture utilisation and sequestration (CCUS). In parallel, CO2 can also be permanently stored in sedimentary formations with poor permeability such as clay- or evaporite-rich deposits, i.e. the opposite of depleted hydrocarbon reservoirs (; Fig. ). The advantage of injecting CO2 in such formations is that no reliable cap rock is needed as in a conventional reservoir, akin to the source rock in unconventional hydrocarbon systems (see Sect. 4.2.1), the impermeable formation will act as both the reservoir and the seal for the injected CO2.

Next to permanent storage in depleted porous media reservoirs and impermeable sedimentary strata, CO2 can also be chemically stored when it reacts with unaltered (fresh) exhumed rocks, a process known as carbonation e.g.. As such, researchers have proposed a link between global phases of plate-convergence-driven mountain-building and cooling climates (e.g. ). Fresh crustal rocks can also be exhumed during plate divergence, e.g. on the rift shoulders of continental rift systems, leading to natural carbonation and storage of atmospheric CO2. Even so, carbonation is much more efficient when (ultra)mafic rocks are available, such as mantle rocks or basalts . Fresh mantle rocks may be exhumed along the rift axis or spreading ridge during the break-up stage in magma-poor systems and during the drifting stage in slow-spreading systems e.g.. CO2 could be artificially injected into these exhumed mantle rocks for storage via enhanced carbonation e.g.. In magma-rich systems, flood basalts that erupted in continental settings, or extensive basalt flows forming the seaward-dipping reflectors during the break-up stage, could similarly be injected with CO2 for permanent storage . The advantage of injection in mafic rocks over porous media reservoirs is the enhanced potential for mineral trapping so that the CO2 will be permanently fixed to the host rock . Projects to explore the possibilities of injecting CO2 into basaltic rocks are under way, for instance, in Iceland Carbfix project; and in the Red Sea, offshore Saudi Arabia .

Finally, the use of nuclear power provides an alternative to burning hydrocarbons and permanently storing the resulting CO2. However, the resulting nuclear waste can remain hazardous for thousands of years, requiring its own permanent geological disposal with isolation and containment demonstrable for up to 1 million years. Therefore, suitably stable, deep (often >200 m b.s.l.), and low-permeability geological formations are required. Salt or clay formations deposited in abandoned rifts are commonly proposed as appropriate geological storage sites due to their extremely low permeabilities and self-healing properties .