By the end of June 2020, the World Health Organisation (WHO) had issued a warning that hospitals were facing a shortage of oxygen concentrators in the face of a Covid-19 pandemic that continued to unfold in different regions around the world.

When Europe and North America were largely the epicentre for the pandemic in the month of March and April, the talk was understandably of the challenges being faced in bulk liquid supply of medical oxygen to hospitals and makeshift field hospitals. The gases industry had risen to the challenge of being on the frontline with remarkable agility and speed, the tales of which were regularly regaled online and in print by gasworld82.

The unsung heroes of this effort must not be forgotten, those who continued to fill and maintain cylinders and cylinder supplies during lockdowns, those who continued to operate oxygen plants, deliver liquid and cylinders to hospitals, maintain hospital systems and maintain supplies to patients in their own homes, whether via liquid, gas cylinders or concentrators.

Yet there were still perceived gaps in the supply chain – such was the level of crisis that healthcare systems were facing. So how is the medical oxygen supply chain structured and where could it have gone wrong? Having explored the discovery and role of oxygen in our previous instalment, here we look at the medical oxygen supply chain and set the scene for the further discussion still to come – our oxygen supply chain preparedness. 

Traditional supply chain

Oxygen in the medical field has traditionally been produced centrally at the ASU(s) and then distributed in liquid form, or as a gas via cylinders, to the customer. 

An alternative and increasingly popular means of oxygen generation exists in the medical market, however, in the form of pressure swing adsorption (PSA). European hospitals have typically relied upon these two choices for supplying medical oxygen to their medical gas network – depending on their consumption they could either purchase liquid oxygen stored in on-site in bulk cryogenic tanks and fed into the facility via its pipelines, or purchase cylinder supply and the regular refills required. 

Bulk supply via storage tanks and pipeline into the wards is generally the most popular or required means of supply, such is the volume of medical oxygen required. However, it is not uncommon to find a combination of bulk and cylinder supply – and especially so during these circumstances of pandemic, whereby the amount of oxygen required could feasibly outstrip the capabilities of the pipelines to deliver it. Cylinder supply can both complement the piped capacity and also provide a relatively mobile means of supply.

Oxygen cylinders come in a variety of sizes to provide continuous or demand flow of oxygen, while alleviating discomfort associated with these conditions. Hospitals often use these cylinders to ensure ease of deployment throughout the hospital, whether at the patient bedside, in operations or procedural areas, in MRI rooms or even integrated into mobile equipment. This flexibility to be used in various medical applications has not only resulted in oxygen cylinders becoming commonplace in the hospital environment, gasworld understands, but had also been crucial in responding to the masses of patients infected with coronavirus at any one time. 

It was commonly reported that hospitals were having to set up makeshift wards in corridors and non-essential rooms, or governments were having to establish improvised facilities in conference buildings to deal with the sheer volume of patients during the height of the pandemic, both of which require the portability of cylinder supply. 

Medical oxygen cylinders are also used by first responders – those paramedics and emergency services personnel that are on the frontline of this crisis. Ambulances, for example, will often use fixed, lightweight oxygen cylinders to service the variety of medical equipment in the vehicle; small medical oxygen cylinders used by emergency personnel are the norm when responding to emergency situations and treating patients. First responders working in makeshift holding wards outside of the hospital or in town and city areas will also be dependent upon these cylinders. Medical oxygen cylinders have been the subject of significant advances through the years, and the healthcare sector has naturally demanded ever-efficient vessels coupled with higher pressures, lower weight and improved delivery systems. Valving systems and associated accessories such as hoses and face-masks have all made major advances too83.

Onsite generation via PSA systems

Back to that third, alternative means of medical oxygen supply – onsite generation. Though commercialised in the 1970s, PSA oxygen concentrators for the supply of medical gas distribution systems have grown on the worldwide market in the last 20 years in particular. As the oxygen is produced onsite, without delivery and storage, medical oxygen generators have convinced many hospitals and healthcare facilities in North America, Africa, Middle East, Asia, and in recent years Europe, that they are able to supply medical oxygen at a competitive price compared to liquid oxygen or cylinders84

PSA systems utilise commonly available components that can greatly reduce the initial capital required compared to the cryogenic production of oxygen, for example, though this must be contextualised by noting that a PSA unit will be installed at one hospital or customer site, whereas the higher capital investment in a cryogenic plant will see that facility serve multiple customers/sites. PSA systems also offer the kind of mobility that address the varying requirements of the hospital and healthcare sector; their rise in this area has been facilitated by ‘monographs’ that deem the use of oxygen in the range of 90-96% purity acceptable. 

Part of this family of independent oxygen generation are oxygen concentrators. The global medical oxygen market has seen the introduction of better, safer and more efficient portable oxygen concentrators in recent years – giving patients with respiratory problems a new lease on life and the freedom to be in their own homes. Portable oxygen concentrators in patietn’s homes have helped to alleviate the pressures experienced by hospitals, managing to free up beds and see more people on an out-patient basis. Even the integration of gas sensor technology into portable oxygen concentrators has assisted in patient care, with individuals better able to monitor their own oxygen levels.


Transporting oxygen via cylinders

As explained in the prior part of this series, for the first half of the 20th century pressurised oxygen therapy was being used to treat numerous medical issues, transported and delivered via heavy steel cylinders. 

An important development occurred in the late 1960s in England and the early 1970s in the US, however, when Luxfer Gas Cylinders introduced seamless, cold-extruded, high-pressure, portable aluminium medical cylinders. Up to 30% lighter in weight than comparable steel cylinders, aluminium cylinders quickly grew in popularity not only among doctors and patients, but also with emergency medical service personnel who benefited greatly from lighter-weight equipment that reduced on-the-job stress85

Today, there are a range of cylinder or packaged oxygen variations to meet the evolved branches of the healthcare infrastructure. Under the homecare offering, an increasing trend for health authorities around the world to enable patients to remain in their home for as long a period as possible, oxygen is delivered in cylinders to a patient’s home to support their condition. As an example, Luxfer makes 300 bar carbon composite cylinders which deliver greater volumes of oxygen in the lightest-weight package available (as light as 0.8kg for a one litre water capacity cylinder) which means patients can remain ambulatory longer, improving their quality of life. Smaller-sized cylinders means easier transport into the household and smaller delivery vehicles too86

In the conventional hospital setting that we are so much more familiar with, it will often be a five litre cylinder we see, each weighing around 5kg. The weight of the 5kg product allows hospital staff to easily deliver product around the hospital and many will be fitted with so-called ‘bed hangers’ that allow the cylinders to be fixed to the patient’s bed. Aluminium alloy cylinders have been developed for use in the hospital setting, which enables them to be used within MRI (magnetic resonance imaging) environments; steel cylinders cannot be used in an MRI room as the strong magnetic fields can cause the cylinders to move, with the potential to harm individuals within the vicinity. Luxfer has also developed these composite cylinders to a higher pressure (300 bar versus traditional 200 bar steel cylinders) to increase the quantity of gas stored. The heavier weight of the cylinder is offset by the ability to hold 50% more gas and incur less refills87

Aforementioned portable oxygen cylinders may also be used as part of mobile equipment in hospitals, often integrated within paediatric incubators. It is understood that these are being integrated into new portable hospital products. 

Of course, to get to the hospital setting, it may well be the case that an individual requires transport via ambulance with first responders. Here there is also a requirement for cylinder oxygen. Ambulances typically have a fixed oxygen cylinder that services the equipment in the vehicle, usually of a portable weight and size that allows them to be swapped over easily. Doctors and paramedics who work from ambulances, helicopters, fire engines and motorbikes will all have small medical oxygen cylinders in their emergency bags. The same requirement for light weight and portability applies to the military sector, where many military medics will carry lightweight oxygen supply in their medical field kit. Cylinders are also provided in survival kits for military jets so that an ejected pilot can treat themselves in hostile environments88

It is clear that since the first recorded discovery of oxygen circa 245 years ago, the vessel upon which it is transported has been both in constant demand and constant evolution. The same is true still today, in 2021, where companies are still looking for even lighter weight and higher pressure cylinders and the need has been not only exacerbated but also greatly highlighted by the shortages in oxygen capacity at the point of use, in higher and lower income countries alike.


Distribution network…

Having understood the means of production and transportation, what of the respective oxygen capacities and infrastructure worldwide? 

It’s difficult to pinpoint an exact installed base of oxygen production capacity, given the many small and large oxygen production plants globally, many of which may not be known or capacities undisclosed, while significant volumes will be captive capacity (privately owned and operated by a steel company, for example) and not merchant (customer) supply. With PSA oxygen generation gradually on the rise, it is also challenging to put estimates beside this varying range of supply capability. There likely exists a very broad sliding scale of oxygen supply capacity at any one time. 

What we do know, however, is both the commonly accepted utilisation rates of ASUs (production plants) and the understanding from within the industry and externally that a large percentage of any bottleneck issues sits within the distribution chain and not in production. 

It’s understood that ASUs typically run at utilisation rates of circa 70-80% of their oxygen capacity, so ramping up capacity within the installed infrastructure should not be a problem in itself, assuming captive plants are willing and able to do so of course. The challenge is to package and transport that increased product, and ensure it reaches its destination and can be appropriately accessed/used once there. 

Karina Kocha, Business Intelligence Manager for gasworld, had explained during the height of the pandemic in Western Europe and North America that this was the case; that it seemed the challenge was not so much in the supply of oxygen in the market, but rather the capacity of the hospital supply systems to cope with the sharp increases in demand as they become overwhelmed with stricken patients. 

“The problem here may be in the capacity of the hospital supply systems, which could be not enough for increased consumption. The fastest way to increase it is to bring oxygen in cylinders…. The amount of oxygen can hardly be a problem. Any shortage might be caused by the delivery system, which can serve as a bottleneck between producers and patients.” 89 

According to gasworld Business Intelligence, quoted on 25th March 2020, the production capacities of liquid oxygen in the US allowed for an estimated production of around 23-26,000 tonnes per day (tpd) of liquid oxygen, with a utilisation rate of 75-80%. As much as 2,600 tpd of oxygen is consumed by the healthcare sector, including hospitals and homecare (in normal conditions). Catering for increased oxygen consumption may require an additional 300-350 tpd of oxygen, which would represent around 12-14% of current medical consumption or 1-2% of liquid oxygen produced 90

“We believe that an additional amount could be obtained through increasing the utilisation rate, as well as the fact that industrial consumption is expected to be decreased to some extent – releasing spare volumes of oxygen for medical usage,” Kocha had affirmed.

When the story of crippling oxygen shortages would be seen in India 12 months later, the message was much the same – in terms of the known installed base of oxygen capacity, and commonly understood utilisation rates, there should have been enough liquid oxygen available to meet spiralling demand.

Asked for comment and clarity, Kocha was again quoted as affirming, “According to our data, the total annual production capacity for oxygen in India is around 65,000 tonnes per day (tpd) for 2020/2021. This is the total amount of oxygen that can be produced at all oxygen facilities in India.” 

“The steel industry requires just over 70% of this volume, another 15% is consumed by the petrochemicals sector, while the non-ferrous industry metals consumes about 5%. The ‘normal’ medical oxygen demand, according to our estimations, was around 1,600 – 1,700 tpd in the period from 2016-2019 and before the Covid pandemic struck. Due to Covid-related oxygen demand that appeared in 2020, however, oxygen consumption in hospitals rocketed to about 10-12,000 tpd in 2020 and is expected to be about 8-10,000 tpd in 2021.” 

“What does this mean? Whilst a theoretical approach,” she added, “it reveals that the main problem is not in a lack of production. When we talk about availability of oxygen, we also talk about the ability to deliver and distribute – to organise logistics, to have enough cryogenic (tank) trucks, cylinders, and refilling facilities. The situation here looks like it is not production, but the distribution system that has become for India the bottleneck on the path of oxygen from factories to hospitals.” 91 

Whilst true, it became clear that the sliding scale of production capacities was not enough to cope with the rapidly escalating demand for oxygen, and often in remote hotspots where the nearest installed oxygen production plants were simply too far away. Previously unheard of scenes followed, as international efforts to ship, fly and truck tanks of liquid oxygen into India ensued. Part of the conundrum of these efforts concerned the science and efficacy of moving molecules of oxygen, given its significant liquid-to-gas expansion ratio. Moving oxygen in its liquid state, via bulk distribution, is considered the most efficient but requires specialist transportation equipment, rendering this a challenge depending on the journey at hand; conversely, if oxygen is transfilled into more mobile cylinders in its (compressed) gaseous state, you then face the inefficiency of shipping or moving a container that weighs more than its contents. In the case of the latter, the necessary equipment is required to facilitate the transition from liquid to gaseous oxygen and complete the cylinder filling process.

How does it look tomorrow?

Whether an accurate estimate of the installed oxygen capacity around the world can ever be achieved, is perhaps a wider question for the industry and world health leaders. It is known that since the first recorded discovery of oxygen around 250 years ago, we have come to rely on this most precious of gases for sustaining life and wellbeing. And yet, it has become clear during the Covid-19 pandemic, if not before, that our traditionally accepted supply chains have not been able to meet the greatest of demands when we needed them most.

Medical oxygen shortages around the world have been a tragic feature of the pandemic, impacting the poorest countries disproportionately. In those countries, where there is not the required access to oxygen or the infrastructure to bridge that gap between supply and demand, the reality is that it is far easier to transport or spready a virus like Covid-19 than it is to distribute the necessary oxygen. These access difficulties were entrenched in many parts of the world before Covid-19, and have been exacerbated by the pandemic, putting strain on fragile health systems and resulting in preventable deaths.

At the time of writing on 29th June (2021), estimates from PATH, a global non-profit organisation for improving public health, suggested that around one million critically-ill Covid-19 patients in low and middle-income countries need two million oxygen cylinders (14.2 million cubic metres) per day92. The question then is, not necessarily how the medical oxygen supply chain is structured today, but how do we need it to look tomorrow? 

How can we learn from the lessons of 2020/21 and improve access to medical oxygen? How can we attempt to keep pace in its distribution? How robust could our oxygen supply chains be, and how pandemic-prepared will we be in the future?