Designing Oxygen Concentrators for Low Resource Settings Part 3: Power Resilience and Energy Efficiency

This is the third part of a 3 part series sharing some of our learns from designing an oxygen concentrator for low resource settings.

Part 1: Designing Oxygen Concentrators for Low Resource Settings: Part 1 Resilience to Dust

Part 2: Designing Oxygen Concentrators for Low Resource Settings: Part 2 Resilience to Moisture

Resilience to poor quality power and grid power failure

The electrical requirements of static oxygen concentrators, such as the AirSep Elite and DeVilbiss 525KS [1], which are frequently used in low resource settings, are a constant 230VAC at 50 Hz. They also consume a lot of energy: existing 10 L/min commercial oxygen concentrators consume around 600W, whatever the output flow rate of oxygen selected by the user.

Power systems are inherently dynamic entities that are expected to fluctuate during normal operation. However, in low-resource settings (LRS) power quality issues can be significant, with outages, voltage sags, surges, spikes and frequency fluctuations not uncommon. In fact, poor quality mains power supply is known to be one of the primary causes of malfunction and failure of medical equipment, especially oxygen concentrators in low- and middle-income countries (LMICs) and has already been discussed by us elsewhere [2] alongside potential solutions which are currently available for existing oxygen concentrators. However, the standard equipment required to protect an AC induction device (like a fridge or oxygen concentrator), includes a heavy transformer and is rather unwieldy (see Figure 15b). After handling this solution in practice, although on one hand it does works effectively and protects the system from electrical faults, on the other hand it does not seem very elegant.

In contrast the concentrator we have been designing works on DC power. This makes power protection somewhat simpler — AC power is converted to DC; a wide range of AD-DC convertors are available with different input ranges and specifications, depending on the requirements of the compressor. For the oxygen concentrator that we have been developing a AC/24V DC converter was selected, which has been designed for the coal mining industry (see Figure 15a). This has an ultra-wide input range of 85–850VAC, will work between -25°C to 70°C. It also has a high isolation test voltage of 4000VAC (and is therefore able to survive significant voltage spikes). The device is rugged and provides the concentrator with a good level of protection against poor mains quality power, while being more cost effective than the transformer based protection for AC driven oxygen concentrator compressors. It also significantly lighter (~0.84kg, compared to 7.2kg for the Sollatek device). The AC-DC converter was housed, along with other power system related electronics, in an IP rated aluminium case within the oxygen concentrator, isolating it from potential damage by dust, if somehow this did manage to enter the internals of the system.

Figure 15: Designing an oxygen concentrator that is resilient to poor quality power supply. Links to Sollatek devices: Voltright SVS04–22E and FSP OE

Managing power outages

There are a number of good solutions available for when the power goes down. For example, FREO2 have developed a low pressure oxygen store (LPOS) designed for maintaining an oxygen supply during blackouts. The storage system saves any excess oxygen that is generated by the concentrator and this is used to deliver a continuous flow to the patient if and when the power goes down [3]. This is a simple and easy to maintain approach and FREO2 have deployed it and are currently testing this in East Africa. Diamedica has also developed an oxygen reservoir able to store up to 500L of oxygen at 5 bar. The solution is commercially available [4]. These are good solutions which can be employed in low resource settings where there are concerns around power supply intermittency.

As well as an 230 VAC input, our oxygen concentrator system includes the ability to power the concentrator from a DC input, using a DC-DC convertor to allow anywhere between 12 and 48V (see Figure 16). This allows the system to be powered from a car cigarette lighter, a 12V battery pack or the output from the charge controller of a larger solar array. The system has an inbuilt UPS, so that if the mains goes off, the alternative power supply would take over. We opted to not include a battery within the unit itself, but instead allowing multiple voltage inputs allows users to make a choice about how much battery storage capacity they want to include and how they charge their storage — e.g. it could be a battery bank charged by solar power.

Figure 16: A DC input between 12–48V can be connected, e.g. this could be a battery charged by a solar array. The system switches to use this power source when there is an interruption in mains power.

More Energy Efficient Oxygen Concentrators

Existing commercial oxygen concentrators consume a lot of energy, with a typical 10 L/min unit requiring around 600W. With nearly a quarter of primary health centres in Sub-Saharan Africa having no access to electricity, and many with very poor reliability, the development of high performing, energy efficient concentrators could increase the viability of producing oxygen in remote health facilities using solar energy. Improved energy efficiency would allow for the use of smaller batteries and solar panels which not only enables appliances to run for longer but also reduces the cost of capital investment. Unlocking a remote clinic’s ability to generate their own oxygen would have very significant improvements on health outcomes.

So how can we improve the energy efficiency of concentrators that have been on the market for over twenty years? Since the first concentrators were designed there have been a number of innovations and advancements in different fields which might make this possible.

Saving Oxygen

Before we actually talk about how we might improve the energy efficiency of an oxygen concentrator it is important to consider how we might:

1. Prevent wastage and make the best use of the oxygen that we do generate. This is because we breathe in for less than half the time that we breathe out, yet oxygen is delivered to a patient continuously. If oxygen were only delivered when the patient breathed in, a significant amount could be saved.
2. Run the oxygen concentrator in the most efficient way possible and store any surplus oxygen for when the power does go down.

Preventing Wastage

Figure 17 shows the respiratory cycle in relation to providing oxygen with a continuous flow system. This shows that the proportion of oxygen wasted is substantial. Saving this oxygen would make a significant difference to the energy efficiency of a system.

Figure 17: The respiratory cycle in relation to providing oxygen.

There are potentially a number of ways to reduce the wastage of oxygen. Three ways that are established and commerical available include:

1. Pulse Flow — Some oxygen concentrators have a system called ‘pulse flow’, where the oxygen outflow is triggered by the patient breathing. Normally this is electronically controlled. Currently these are typically designed to work with adults, however, it was suggested in an Oxygen Colab workshop discussion that a version could be designed to work for babies. There is potential for this to be designed as an ‘add on’ device that could be used with any oxygen concentrator system.
2. Nasal Reservoir — The other simpler approach discussed in one of the Oxygen Colab workshops was the use of a nasal reservoir. This is known as a non-rebreather mask, where the mask is connected to a plastic reservoir bag which fills with oxygen from the concentrator. The mask has a one-way valve system that prevents exhaled oxygen from mixing with the oxygen in the reservoir bag. (This is different from a partial rebreather mask, which looks similar to a non-rebreather mask but contains a two-way valve between the mask and reservoir bag). The concentrator fills the reservoir bag at a constant rate, inflating as the person breathes out, therefore requiring a lower flow rate than is required with a nasal cannula.
3. Oxygen Conserving Reservoir Devices — These work by pooling a small reservoir of oxygen (~20ml) close to the airway. They are normally in the form of a moustache-reservoir that sits under the nose, or a pennant-type device that hangs around the neck. These devices are simple and have been proven in clinical trials to be effective [5].

Converting the system to run on BLDC

BLDC motors provide a robust way to directly improve the energy efficiency and flexibility of oxygen concentrators for low resource settings. They allow the power consumption of a concentrator to scale with the oxygen flow rate, making it possible to supply oxygen with less capital expenditure in off-grid locations. When less power is available, the flow and/or concentration can be reduced (to a minimum of 82%) to allow the oxygen to flow for extended periods of time.

BLDC motors would work well with DC solar installations — where no conversion from AC is necessary. They should be considered as part of the solution in designing more resilient and lower power consuming oxygen concentrators, specifically important factors for low resource settings. With nearly a quarter of primary health centres in Sub-Saharan Africa having no access to electricity, and many with very poor reliability (as highlighted previously), high performing, BLDC motors in oxygen concentrators would increase the viability of producing oxygen in remote health facilities using solar energy.

A report published by Efficiency for Access found that permanent magnet brushless motors use between 22–42% less energy than conventional alternating current (AC) motors [6]. Therefore adapting an existing appliance to use a BLDC motor can lead to considerable energy savings.

Lab testing that we carried out does indeed suggest that BLDC motors do save energy compared to their AC counterparts — this can be between 10–25% in energy savings when a direct replacement is carried out — this brings the watts per litre down from 60Watts/Litre to between 45–55 Watts/Litre depending on the set up of the system. This is an important saving over the life of the concentrator and there is scope for further optimisation. However, the greatest benefit of BLDC motors is being able to scale the power in proportion to the flow of oxygen required. So for example, at a flow of 2 Litres/min, a BLDC motor can be adjusted to consume only around 92W, while still providing 90% oxygen, compared to a 5L oxygen concentrator using an AC compressor, which will still be consuming close to 300 W.

We are about to publish a much more detailed article on this topic.

Using VPSA to improve the efficiency of Oxygen Concentrators

The majority of stationary oxygen concentrators on the market use pressure swing adsorption to concentrate oxygen. Each sieve bed is alternatively pressurised, concentrating oxygen before being depressurised to atmospheric pressure, in a cycle. With VPSA, after depressurising the sieve bed, it is pulled down to a vacuum.

There are a few potential benefits of using VPSA. Firstly because the system is operating at a lower operating pressure, the energy consumption is lower. Secondly, the low pressures reduce the potential for water condensation — this means VPSA systems are theoretically less susceptible to humid environments compared to PSA systems. Thirdly, the lower operating pressures lead to reduced attrition within the zeolite beds and therefore less zeolite dust and longer bed life. These reasons combine to give longer life sieve beds — reducing the lifetime cost of the system.

One of the challenges of using VPSA within a stationary oxygen concentrator is the need to have both a compressor and a vacuum pump. To get around this, we explored converting an existing dual head compressor to be able to compress and draw a vacuum at the same time; this only requires some small modifications to the one way valves within the compressor as well as preventing air flow between the two heads of the compressor. Although this is by no means an optimal way to achieve this (i.e. the ratio of pressure to vacuum could not be controlled), it was effective and proved the concept. After discussions with the manufacturer of the compressor they indicated that they were open to making these modifications within the manufacturing process — it would not be a significant change to the existing manufacturing process. We adapted the Devilbiss rotary valve system to allow us to be able to alter the timings and operate a VPSA system. We are also in the process of designing a new rotary valve system specifically designed to work with VPSA systems. In our testing we were able to obtain around 45 Watts per litres of oxygen compared to 60 Watts per litre with conventional systems. Further improvements are possible by optimising cycle timing and equalisations, and this initial exploration has given us confidence that a VPSA system has potential to be both practical and beneficial. We continue to do further work in this area and will share our findings in future article.

Sieve Bed Innovations

There is scope for innovation of the sieve bed materials and structure. This has not been explored in our scoping work, however, some things which should be considered include:

1. Enhancing Zeolite — Zeolite molecular sieves have been used in industrial applications for over 65 years, mainly for gas separation. There has been considerable innovation in enhancing zeolite in this period, including enhancing nitrogen capacity, selectivity and mechanical durability. Further innovation is possible, however, with existing zeolite materials (e.g. lithium, sodium enhanced 13X type zeolites). This is more likely to be an incremental improvement rather than a significant step change in performance. For example see the work of TDA research’s work on a modified version of a the lithium exchanged low silica A zeolite where they were able to enhance nitrogen adsorption capacity [7].
2. Structured Beds — Zeolite molecular sieves in oxygen concentrators are typically used in a bead form. However, potential improvements through forming compact structures such as honey combs, and multi-channel or foam type structures. By choosing appropriate geometries, the packing density can be increased without increasing the pressure drop. Furthermore, structures can be designed to guide the gas flow through the structure to ensure adsorption is optimised.
3. Metal Organic Frameworks (MOF) — Step change improvements, however, might be expected from advancements in metal organic frameworks (MOFs). These consist of metal ions connected with organic ligands that are arranged in a structure framework. MOFs are more tunable than zeolites in terms of designing them to have desired adsorption properties: e.g. strong binding forces between MOF structure and N2 compared to O2, while ensuring easy desorption of the nitrogen from the material. Furthermore, higher uptake capacity means less cycles for the same amount of oxygen and will subsequently reduce the required energy for the concentrator.

Potentially Disruptive Technologies

There are a number of emerging technologies, which might in the future be more appropriate for generating oxygen in remote health care centres and which use less energy for litre/min of oxygen. There is some cross-over here with the previous article of resilience to moisture.

1. Paramagnetic oxygen separation — This is where a strong magnetic field is used to do the gas separation. This is potentially game changing, as the energy budget would be drastically reduced, and there would be less parts (such as the sieve bed) which would be susceptible to the kind of damage that causes existing concentrators to fail [8].
2. Using algae to generate oxygen — It is possible to produce large volumes of oxygen from algae through photosynthesis, a process which can be run directly from sunlight or from LED light sources. Algae consumes CO2 and releases oxygen so it is ideal if it can be linked with a process that is producing CO2, such as brewing, or combustion. It was also mentioned in a Oxygen Colab workshop that algae could be used to scrub CO2, e.g. scrubbing CO2 from the exhaled air of a patient, and recycling this to reduce the oxygen concentration demand. This would lead to a massive overall gain in system efficiency — because overall less oxygen has to be produced to provide to the patient. This technology is well developed and NASA has an algae bioreactor on the International Space Station which is scrubbing CO2 and providing O2 for crew. It could run directly from sunlight during the day, with nighttime O2 being supplied by O2 storage or solar/battery storage running LEDs in the bioreactor.
3. Four Stroke Direct Compression VPSA — The PSA process could be enhanced through a four stroke direct compression VPSA based design. The energy of the compressed nitrogen is lost in most ‘two stroke’ compressors. In contrast, 4-stroke compressors synchronise adsorption and desorption so that the system captures the energy of expansion of the waste N2.
4. Free Piston Stirling Engine — Running an oxygen concentrator on a free piston stirling engine, either using the heat direct from sunlight or another waste heat source. This is already done for camping fridges, for example, by a company called ‘Stirling Ultra Cold’ [9]. Stirling engines are also used in satellite power generation and submarine propulsion. This approach seeks to run the compression cycle without electricity. There is currently nothing off-the-shelf that would fulfil the requirements to run an oxygen concentrator, however, there is perceived potential in this technology. The piston-free stirling engine has only one moving part and no seals to wear. Night storage could be provided through thermal storage solutions, which are well established such as molten salt storage.
5. Ceramic Oxygen Generator — The ceramic oxygen concentrator has been developed by NASA and other research partners. It is a solid state device with no moving parts that is durable. It uses a type of ‘wafer’ that when heated, pulls oxygen out of the air. Ceramic membranes have been intensively studied in the past two decades, by researchers in Imperial College London, Nanjing Tech University, and Dalian Institute of Chemical Physics. As well as NASA, Nanjing Tech University has also developed a prototype of a ceramic membrane based oxygen concentrator [10]. Ceramic oxygen concentrators would consume significantly less energy than existing oxygen concentrator devices.

Conclusions

Improving the power resilience and energy efficiency of oxygen concentrators is an important step towards creating concentrators that are fit for low resource settings. Benefits include the ability to power the concentrator in regions with poor quality power supply as well as with smaller and more affordable batteries, solar panels or other renewable sources allowing remote off-grid (or weak grid) clinics to generate their own oxygen; enabling concentrators to run for longer on a given size of battery (which may be vital for continuous production of oxygen throughout the night). There are various mature technologies, such as BLDC motors and power protection technology, that are available that can be used in today’s concentrators to achieve this with very little modification. There are also established ways of saving oxygen, which may well be the most cost effective approach. Furthermore, there are also a number of game changing technologies that have the potential to revolutionise the process of oxygen generation altogether.

This is the final part in a three part series sharing some of the lessons we have learnt while designing an oxygen concentrator that is more resilient in low resource settings.

We’d love to hear what you are working on and ideas of how to reduce energy consumption in oxygen concentrators. Do reach out and get in touch.

References and Notes

1. https://www.drivedevilbiss-int.com/products/respiratory/oxygen-therapy/149/compact-525?number=525KS
2. Power Protection for Oxygen Concentrators: the why, what and how, Published by COVIDaction through Medium, 2019, Joel Chaney
3. https://freo2.org/products-innovations
4. http://www.diamedica.co.uk/english/product_details.cfm?id=1565
5. An example of one of these devices is the Oxymizer (see https://pubmed.ncbi.nlm.nih.gov/2924615/)
6. The Benefits of Permanent Magnet Motors: Efficiency Opportunities in Off and Weak Grid Markets, Low Energy Inclusive Appliances, Efficiency for Access, 2020 https://storage.googleapis.com/e4a-website-assets/The-Benefits-of-Permanent-Magnet-Motors_Efficiency-for-Access.pdf
7. A Low-power Medical Oxygen Concentrator G. Alptekin, D. et al., 46th International Conference on Environmental Systems, https://ttu-ir.tdl.org/bitstream/handle/2346/67687/ICES_2016_360.pdf?sequence=1
8. See “Cost-Effective Oxygen Separation System Based on Open Gradient Magnetic Field by Polymer Beads”, Dr Raghuvir Singh et al, https://netl.doe.gov/sites/default/files/event-proceedings/2015/gas-ccbtl-proceedings/DOE-workshop.pdf
9. https://www.stirlingultracold.com/stirling-engine/
10. See https://www.sciencedirect.com/science/article/pii/S1000936116300589 and https://core.ac.uk/download/pdf/268871461.pdf for some more details