The Linde-Evonik pilot & its implications for the Europan Hydrogen Backbone

Unlike the US, Europe has been quite ambitious & pro-active about climate change initiatives at scale. In line with the commitments made at the COP in Paris, the EU aims to become climate neutral by 2050. An important milestone in that journey is to achieve 55% reduction in emissions by 2030!

When we talk about climate change, there are essentially two legs – a cleaner fuel which doesnot add to the GHG load & a way to capture those that are being emitted or are already present in the atmosphere. While a lot of routes to capture and use carbon dioxide are being explored, as far as fuels are concerned, hydrogen is pretty much the automatic choice. But when it comes to hydrogen, challenges lie elsewhere. The flammability risk coupled with difficulty of detection makes it risky to transport as a gas & unlike gaseous fuels that we are used to dealing with (LPG & natural gas), hydrogen is extremely expensive to liquefy.

To liquefy gases, we may need to reduce temperature, increase pressure or, in some cases do both – depending upon the critical temperature of the gas vis-à-vis ambient. To be able to liquefy a gas only by applying pressure, it must be maintained below its critical temperature. Natural gas is mostly butane which has a critical temperature of 96C, so it can be conveniently liquefied at ambient temperature. The critical temperature of hydrogen on? the other hand is as low as -240C which means that to liquefy hydrogen we have to first cool the gas down to that level and, once liquefied, store it in insulated containers at high pressure – about 42 atmospheres to prevent it from evaporating – all of which makes it uneconomical to scale-up.

Another option is to convert hydrogen to a compound like ammonia which can be liquefied and transported. This however means two additional steps – converting hydrogen into ammonia at the point of manufacture (or starting with ammonia itself) and at the receiving point, splitting ammonia to its components: hydrogen & nitrogen and separating them as pure streams – which means additional equipment, operating cost and losses. Options like replacing 10% of the natural gas with hydrogen in building heating are being actively pursed where the cost of moving to a cleaner technology represents an incremental increase - but all these options are good only as short term measures. To make hydrogen central to the energy economy, inevitably a cost-effective solution based on pipeline transport has to be found.

Recognising the challenge, way back in July 2020, a group of European energy infrastructure operators came together to form a consortium called The European Hydrogen Backbone (EHB) initiative. It envisioned a dedicated hydrogen pipeline infrastructure, to a large extent based on repurposed natural gas pipelines, to transport hydrogen at high pressure across the member countries. It has since grown to 31 European network operators with infrastructure covering 25 EU Member States plus Norway, the United Kingdom, and Switzerland.

The invasion of Ukraine by Russia, however, put a completely different spin to this. Pivoting from the earlier plan that was centered around climate change, the EU leadership decided to make hydrogen central to their energy resilience strategy with a view to significantly reduce imports of Russian gas by 2030. This, however, means that they now need to quadruple the planned capacity of the hydrogen supply chain from 5.6 million tons to 20 plus million tons – all in the next 6 years! Initially, a third of the additional 15 million tons would be locally produced and the rest imported. Suddenly, finding a cost-effective transport solution for hydrogen is no longer just good to do, it is strategic for the energy security of the continent.

Linde, a world leader in gas separation & Evonik, a specialty polymer manufacturer have tied up to offer an end to end solution to separate 99.9999% pure hydrogen from a feed where the balance 95% is natural gas. If successful, it will bring down the cost of transporting hydrogen significantly because it can travel as a mixture with natural gas through the current infrastructure, while giving the flexibility to extract pure hydrogen at the receiving end, anywhere in the network! The solution is based on two Unit operations in series - membrane based separation & pressure swing absorption & we will see later how they complement each other.

Use of membranes for gas separation has been of interest since the 1960s since energy consumption is low but the challenge has been mainly around the degree of purity achieved. Studies on gas separation indicate that the efficiency of gas transport is optimal when one of the gases has significant affinity for the membrane polymer as well as a high rate of diffusion through it. Diffusion is a function of size, while membrane affinity is influenced by distribution of electrical charge within the molecule & often the two properties are not in sync. For example, if we try separating a mixture of CO2 & H2 using a cellulose acetate membrane, hydrogen on account of its smaller molecular diameter diffuses much faster but since it cannot bind to the membrane as effectively as CO2, separation is suboptimal. This is not unique to CO2-H2, it is infact, common for many pairs of gases and the level of separation possible is referred to as Robeson Upper Limit. To achieve a minimum purity of 70% to feed into a PSA process for hydrogen in this case, Evonik needed to find a way to beat this limitation. The schematic given below broadly explains how it works.

The membrane is in the form of a bundle of long narrow tubes, placed lengthwise in an airtight housing. The open end of the tubes is connected to a vacuum system to create a pressure differential between the outer and inner surfaces of the membrane. The gas mixture is introduced at one end in the space between the tubes and the housing. The hydrogen molecules permeate into the tubes from where they are sucked out. The bulk of the natural gas molecules leave the other end of the housing, now practically free of hydrogen.

The high permeability of hydrogen, complemented by the pressure differential between the inside and outside surfaces, ensures yields of about 90% purity – which feeds into the PSA module.

You may recollect Pressure swing adsorption from the Covid days when it became the mainstay to address the oxygen crisis. It is a molecular level purification process where out of the mixture of gases fed into a packed bed & some are selectively retained on the surface, leading to separation of the desired component up to a very high level of purity. The extent varies, but in general, adsorption is proportional to the operating pressure, which means by changing the pressure the amount of adsorption can be varied. The pressure in the absorption column swings alternately between high & low and hence the name Pressure Swing Adsorption.

The schematic below pictorially depicts how a PSA system works. The adsorbents used in PSA process are often referred to as molecular sieves and it is a nice way of visualizing what they do. The molecular sieves have apertures which are angstrom level in diameter (1 angstrom is 1 millionth of a mm) and can be designed for the combination of gases that they need to separate. The smaller particles pass through while the larger ones block the pores. As the bed gets saturated, the rate of adsorption drops & it needs to be regenerated. To do so, all that needs to be done is to de-pressurise the bed for the adsorbed gases to be released from the surface and then it is ready for the next cycle again. Usually commercial installations have multiple adsorbers in series and while one bed is under adsorption, the other is undergoing regeneration, ensuring a steady flow of gas. Starting with 90% purity input, it delivers hydrogen of purity exceeding 99.9999%.

Just an interesting aside: when the PSA plants were set up to address the oxygen crisis at the peak of the Covid pandemic, the job was to separate oxygen – which constitutes 21% of air from nitrogen which is about 78% - which at first glance doesnot seem very far from what we were trying to achieve with hydrogen: but there is a catch. The purity of oxygen required was around 92% and this could be delivered through a PSA system using ambient air as feed. On the other hand, the purity required for hydrogen was 99.9999%. Expressed in terms of impurity as parts per million, for the hydrogen it is 1 part per million, whereas the oxygen stream tolerated 80,000 parts per million – so the challenge was significantly higher.

The Linde-Evonik solution is unique in the way two technology leaders have come together to provide an end-to-end solution. Bridge technologies such as these which harness existing knowledge & capabilities to offer new solutions, play a critical role in managing transitions by managing risk, while maintaining the momentum.