Greenhouse gas (GHG) emissions, of which the healthcare sector generates 4% to 5%, are causing a rapidly increasing global surface temperature. This leads to climate change, which is one of the fundamental threats to human health. We urgently need to reduce our GHG emissions, but to do so efficiently, we need to understand which activities contribute to the highest emissions. Life cycle assessments (LCA) play an important role in this understanding.
LCA is a systematic analysis of the environmental impact of a product or process from its material origin to its end-of-life. LCA data is fundamental for identifying hotspots, which are steps and inputs most responsible for negative environmental impact, and for providing the focus for action.
Here, we compare the environmental impact concerning carbon footprint from a legacy protein A downstream purification process using rProtein A Sepharose™ Fast Flow (FF) chromatography resin to the more modern processes using MabSelect SuRe™ resin and the most recently launched MabSelect PrismA™ resin. In cradle-to-gate LCA, the carbon footprint per liter of resin associated with chromatography resin manufacturing is the smallest for the legacy resin. However, MabSelect PrismA™ resin has the lowest aggregated carbon footprint in cradle-to-grave LCA, where we are including the entire life cycle of the resins and normalizing the environmental impact to the performance of the resins. The high mAb capacity and long resin lifetime of the MabSelect PrismA™ resin improves the environmental sustainability of the protein A process in high-demand mAb manufacturing where the full lifetime of the resin is utilized.
With a US average electricity grid mix in the use phase, the carbon footprint associated with the energy for operating the cleanroom is the hotspot for all protein A processes. The equipment footprint, cleanroom area, and consequently the energy for the cleanroom infrastructure are the smallest for the MabSelect PrismA™ resin purification process.
Watch this video to learn more about LCA of the protein A process in mAb manufacturing
Introduction
GHG emissions and climate change
GHG emissions are leading to global warming and the global surface temperature is increasing rapidly as seen in Figure 1 (1). According to the Paris Agreement (2016), the global temperature rise this century should be kept well below 2°C above pre-industrial levels and limit the temperature increase to 1.5°C above pre-industrial levels. Climate change caused by global warming leads to irreversible biodiversity loss and extreme weather events such as drought, flooding, and wildfires. Climate change is one of the fundamental threats to global health, negatively impacting mental health and causing malnutrition, allergies, cardiovascular, and respiratory injuries as well as poisoning. Therefore, limiting GHG emissions and the global temperature rise is urgent.
In the 28th annual United Nations (UN) climate change conference (COP28) in Dubai in 2023, an explicit reference was made to the need to reduce 43% of GHG emissions by 2030 and 60% by 2035 relative to the 2019 level (2).
The healthcare sector, which is responsible for improving people’s health, generates 4% to 5% of the global emissions of which more than 50% is driven by supply chain and where 23% to 35% originates from drugs (3, 4). The United States FDA environmental impact review requires that environmental assessments must be submitted as part of new drug applications, and the European Medicines Agency has separate guidance on environmental risk assessment.
For successful decarbonization, we need to understand which activities are contributing the most to negative environmental impact.
Fig 1. Global land-ocean temperature index.
Life cycle assessment
LCA is a systematic analysis of the environmental impact of a product or process from its material origin to its end-of-life deposition. In more detail, a life cycle assessment is an analysis of how a given product or activity affects environmental quality through all stages of its useful life. The analysis starts with the impact on natural resources in obtaining raw materials, through to the product’s manufacture and distribution, moving on through its use or deployment, and ending with the environmental impact of retiring the product and treating its material components as part of the waste stream. LCA data is fundamental for identifying hotspots, which are steps and inputs most responsible for negative environmental impact, and for providing a focus for action (5).
The product in this LCA is the protein A chromatography resin used in the capture step in mAb manufacturing. The cradle-to-gate LCA includes the environmental impact associated with raw materials (including recycling of chemicals), water, energy, and waste in the manufacturing of the chromatography resin (Fig 2). Cradle-to-grave LCA covers the entire life cycle of the protein A resins, also including distribution to the end-user, resin usage in the mAb capture step, and end-of-life management. Note that the remaining process steps, both upstream and downstream within the biopharmaceutical manufacturing process, have not been included in this work.
For the impact of distribution to the end-user, chromatography resin shipped from Sweden to the US was assumed.
In the use phase, we distinguish between chemicals and water for buffers, cleaning-in-place, and storage solution for resins, as well as solutions for cleaning and sanitization of hardware. We also include energy for operating the chromatography systems and the cleanroom infrastructure (heating, ventilation, and air conditioning [HVAC]). For the electricity in the use phase, we assume a US average electricity grid mix but we also compare the impact of alternative energy sources. Column packing and cleaning and sanitization of the chromatography system is not included. For more details see Data inventory protein A capture step (use phase).
End-of-life, that is, disposal of the protein A chromatography resins is incineration of hazardous waste.
Disposal of storage solution (i.e., ethanol and water) is assumed to be 100% municipal water treatment. Disposal of remaining material is modeled as packaging waste. Disposal of packaging is expected to be 100% municipal incineration.
Fig 2. A schematic view of cradle-to-gate and cradle-to-grave LCA for a chromatography resin used for mAb manufacturing.
The data inventory from all activities in the life cycle is translated into environmental impact by using accepted and well-recognized databases. Here, we focus on climate change, which is the same as global warming potential (GWP), that is, all the greenhouse gasses CO2-equivalents (eq), also known as carbon footprint. The IPCC 2021 GWP 100 years method was used for calculating carbon footprint. We also consider direct freshwater use in the protein A chromatography process.
The functional unit in the cradle-to-gate assessment is protein A chromatography bulk resin of a volume corresponding to 1 L of packed resin, determined by the compression factor (see Data inventory). However, 1 kg of mAb is used as the reference flow when covering the entire life cycle (cradle-to-grave) of the protein A resin. The environmental impact normalized to the amount of mAb produced is more relevant for the end-user.
At Cytiva, we generate LCA data according to ISO 14040 to 14044 standards and use a way of working that has been verified to comply with said ISO standards by a third party. SimaPro v9.5.0.0 and the ecoinvent v3.9.1 database were used for background calculations.
Results
Resin manufacturing, distribution, and end-of-life related carbon footprint/L of packed resin
The carbon footprint/L of packed resin associated with resin manufacturing is smaller for the legacy protein A resin, rProtein A Sepharose™ FF resin (launched 1996) compared to the more modern MabSelect SuRe™ resin (launched 2005) and the most recently launched MabSelect PrismA™ resin (2017). However, it is important to understand that 1 L of a legacy chromatography resin cannot be directly compared to 1 L of a modern chromatography resin because of the different performance of the resins.
The carbon footprint from raw materials and emissions associated with protein A resin manufacturing activities (green in Fig 3) is much larger compared to the carbon footprint from the distribution to the end-user (customer), end-of-life, and the plastic container in which the resin is shipped.
Fig 3. Carbon footprint split (kg CO2-eq)/L of packed resin.
Resin, buffer, CIP, and storage solution carbon footprint/L of packed resin and lifetime
The buffers and buffer volumes expressed as column volumes (CV)/L of packed resin are assumed to be the same for all three protein A resins.
MabSelect SuRe™ and MabSelect PrismA™ resins have protein A-derived ligands that have been genetically engineered for improved alkaline stability. MabSelect SuRe™ resin has good alkaline stability and can be cleaned with 0.1 M of NaOH in every purification cycle. MabSelect PrismA resin has excellent alkaline stability and can tolerate 0.5 M of NaOH for CIP in every cycle. rProtein A Sepharose™ FF resin, which is based on recombinant protein A, is not alkaline stable. For rProtein A Sepharose™ FF resin we are assuming a three-step cleaning procedure using 6 M guanidinium hydrochloride (Gua-HCl), detergent (0.5% Tween), and solvent (30% isopropanol), which is performed after each batch (every third cycle). After each batch, which is 3 cycles for rProtein A Sepharose™ FF resin and MabSelect PrismA™ and 5 cycles for MabSelect SuRe™, the column is washed into 3 CV of storage solution (20% ethanol).
In Figure 4, the carbon footprint related to the buffers and CIP solution in the use phase across the entire lifetime of the resins has been added (different shades of brown and orange). The assumption is a high-demand mAb production process (50 kg batch) where the full lifetime of the resin is utilized, which is 100 cycles for rProtein A Sepharose FF resin, 150 cycles for MabSelect SuRe™ resin, and 200 cycles for MabSelect PrismA™ resin.
The increasing carbon footprint contribution from the buffer (brown) going from rProtein A Sepharose™ FF resin to MabSelect SuRe™ resin and finally to MabSelect PrismA™ resin is due to the increasing resin lifetime.
The carbon footprint contribution from the CIP is largest for rProtein A Sepharose™ FF resin even though its resin lifetime is shortest, and CIP is only performed one time per batch, that is, 33 times during this resin’s lifetime. A three-step CIP procedure with Gua-HCl, detergent, and solvent one time per batch results in a much higher carbon footprint contribution compared to CIP with NaOH for MabSelect SuRe™ and MabSelect PrismA™ resins in every cycle. The differences in contribution from the storage solution between the resins are related to the different resin lifetimes and cycling strategies. With 3 cycles/batch for rProtein A Sepharose™ FF resin, the column is washed into storage solution 33 times during the lifetime whereas the same cycling strategy for MabSelect PrismA™ resin results in storage 66 times. For MabSelect SuRe™ resin, with 5 cycles/batch, the column is stored in 20% ethanol 30 times.
Fig 4. Resin (green), buffer/CIP, and storage solution (different shades of brown or orange) related carbon footprint split (kg CO2-eq)/L resin.
Protein A performance: amount of mAb produced/L of resin and lifetime
Both the dynamic binding capacity (DBC) and alkaline stability are improved over the protein A resin generations. The DBC, process load as well as the amount of mAb produced/L of resin and maximum lifetime is presented in Table 1 for rProtein A Sepharose™ FF, MabSelect SuRe™, and MabSelect PrismA™ resins. The process load was assumed to be 80% of the QB10% value and the step yield was assumed to be 97%.
Table 1. DBC, process load, and amount of mAb produced/L packed resin and lifetime
| Resin | rProtein A Sepharose™ FF resin | MabSelect SuRe™ resin | MabSelect PrismA™ resin |
| Dynamic binding capacity QB10% (mg/mL)* | 35 | 44 | 76 |
| Process load (g/L) packed resin | 28 | 35 | 60.6 |
| Alkaline stability | Poor | Good | Excellent |
| Resin lifetime (cycles) | 100 | 150 | 200 |
| Amount of mAb/L packed resin (g/L/lifetime) | 2.7 | 5.1 | 11.8 |
*The residence time (RT) during sample load was 8 min for rProtein A Sepharose™ FF resin, 4 min for MabSelect SuRe™ resin, and 6 min for MabSelect PrismA™ resin.
Carbon footprint split normalized to resin performance: 1 kg of mAb as reference flow
Instead of looking at the carbon footprint/L of resin, it is more relevant to normalize the data to the performance of the resins. We normalized values by dividing the data in Figure 4 by the mAb amounts produced/L of resin and lifetime (Table 1). After normalization and using 1 kg of mAb as the reference flow, the ranking of the resins switches in favor of MabSelect PrismA™ resin. Thus, MabSelect PrismA™ resin has the lowest aggregated carbon footprint/kg of mAb (Fig 5).
Fig 5. Kilogram of CO2-eq/kg of purified mAb under resin lifetime.
The use phase is divided into carbon footprint associated with buffers, CIP solutions, and storage solution (20% ethanol), as in the previous section. Cleaning of buffer tanks and cleaning and sanitization of tanks for the product pool have also been included in this section.
The CIP-related emissions are where we see the biggest difference between the resins. The three-step CIP procedure for rProtein A Sepharose™ FF resin, even though it is only performed after each batch, gives a much higher footprint compared to using NaOH in every cycle for MabSelect SuRe™ and MabSelect PrismA™ resins.
The buffers and buffer volumes in CV are the same for all three protein A resins. However, with the higher capacity resins, the buffer consumption and buffer related carbon footprint/kg of mAb decreases as seen for MabSelect SuRe™ and MabSelect PrismA™ resins. Storage solution (20% ethanol) also contributes significantly. The carbon footprint from the storage solution decreases with increased number of cycles/batch and/or higher capacity as for MabSelect SuRe™ and MabSelect PrismA™ resins.
Cleaning of buffer tanks is only performed using water for injection (WFI). Therefore, the carbon footprint from this activity is much smaller compared to cleaning and sanitization of the tanks for the product pool, which is performed using caustic, acidic, and water of two different qualities.
The carbon footprint/L of packed resin related to the plastic container, distribution, and end-of-life is assumed to be approximately the same for all three chromatography resins. When using 1 kg of mAb as the functional unit as in Figure 5, the contribution from the plastic container, distribution, and end-of-life is inversely proportional to the amount of mAb produced/L of resin and lifetime. We produced 2.7, 5.1, and 11.8 kg of mAb/L of packed rProtein A Sepharose™ FF, MabSelect SuRe™, and MabSelect PrismA™ resins, respectively. Consequently, the contributions from the container, distribution, and end-of-life are almost twice as high for rProtein A Sepharose™ FF resin compared to the MabSelect SuRe™ resin, and more than four times higher compared to the MabSelect PrismA™ resin.
Direct freshwater use/kg of mAb
The direct freshwater use for buffers, CIP, and storage solution for resins, as well as cleaning and sanitization of tanks has been summed up and normalized to the amount of mAb produced (Fig 6). Water use is independent of the resin lifetime but closely related to the capacity of the resins. By using a high-capacity protein A resin like MabSelect PrismA™ resin, the water use will only be approximately 50% of the water use for rProtein A Sepharose™ FF resin. The cycling strategy can also affect the water use for storage solution and cleaning of tanks for the product pool. More cycles/batch for MabSelect SuRe™ resin resulted in a relatively smaller water use for storage solution but higher water use for cleaning of tanks for the product pools.
Fig 6. Direct freshwater use (L/kg mAb) for buffers, CIP, storage solution, and cleaning of tanks.
Carbon footprint including cleanroom and system energy consumption
We estimated the cleanroom area based on the equipment footprint, that is the footprint of buffer-, cell culture supernatant-, and product pool tanks as well as the chromatography system and columns. For all three processes, a BioProcess™ modular system 1½ inch can be used, which has a footprint of 3 m2. The rProtein A Sepharose™ FF resin process requires a close to 600 L column and here, an AxiChrom™ 2000 column is suitable. For MabSelect SuRe™ and MabSelect PrismA™ resin with column volumes of 286 and 275 L, respectively, AxiChrom™ 1400 columns are suitable. The total equipment footprint for rProtein A Sepharose™ FF and MabSelect SuRe™ is approximately 40 m2 whereas the equipment footprint for MabSelect PrismA™ resin is only close to 30 m2. The difference between the different resins and processes is primarily due to the larger buffer tanks for MabSelect SuRe™ and rProtein A Sepharose™ FF resins compared to MabSelect PrismA™ resin. The cleanroom area is assumed to be 10 times the equipment footprint. This results in cleanroom areas for the rProtein A Sepharose™ FF and MabSelect SuRe™ resin processes of approximately 400 m2 and for the MabSelect PrismA™ resin process of close to 300 m2 (Table 2).
Table 2. Equipment footprint summary and cleanroom size for high-demand processes: 50 kg mAb batch
| Protein A resin | Buffer and cell culture supernatant tank (m2) | System (m2) | Column (m2) | Equipment (m2) | Cleanroom area (m2) |
| rProtein A Sepharose™ FF resin | 31.5 | 3 | 6 | 40.5 | 405 |
| MabSelect SuRe™ resin | 31.5 | 3 | 4 | 38.5 | 385 |
| MabSelect PrismA™ resin | 22.0 | 3 | 4 | 29.0 | 290 |
Figure 7 shows the carbon footprint in kg CO2-eq/kg mAb associated with resin manufacturing, distribution, use phase (the mAb capture step), as well as the end-of-life management. The usage now also includes the energy for operating the chromatography system and cleanroom infrastructure, which are new additions in this section. Buffer, CIP, storage solution, and cleaning of tanks have also been included in previous sections.
The carbon footprint related to the cleanroom (HVAC) energy represents the hotspot in the mAb capture protein A step for all three processes when using a US average electricity grid mix. The equipment footprint, cleanroom area, and consequently the energy for the cleanroom infrastructure is smallest for the MabSelect PrismA™ resin process. The carbon footprint associated with the energy for operating the chromatography system is very small. Carbon footprint associated with manufacturing of the hardware (chromatography systems and tanks) is considered as production capital goods with long lifetime and therefore not included in the model for simplified analysis.
Fig 7. Carbon footprint across the entire life cycle of rProtein A Sepharose™ FF, MabSelect SuRe™, and MabSelect PrismA™ resins including energy for operating chromatography system and cleanroom infrastructure, assuming a US average electricity mix.
Figure 8 shows the carbon footprint difference using selected country electrical grid mixes as well as renewable (wind certified) energy for operating the cleanroom infrastructure and equipment in the use phase for the MabSelect PrismA™ resin process. US average electricity grid mix presented in Figure 7 was assigned the 100% carbon footprint. France represents predominantly nuclear power which results in a 68% reduction of the carbon footprint whereas China represents predominantly coal generated power with an 83% increase of carbon footprint. Wind-certified power represents a renewable electricity source resulting in more than 80% reduction of the carbon emissions compared to the US average electricity mix. Transition away from fossil fuels is in line with the outcome from the United Nations Climate Change Conference (COP28) in Dubai 2023 (2).
Fig 8. Carbon footprint (kg CO2-eq/kg purified mAb) differences using selected country electrical grid mixes and wind certified energy for the MabSelect PrismA™ process. “US” is US average electricity grid mix, “France” represents predominantly nuclear power and “China” represents predominantly coal generated power. Wind certified power represents a renewable electricity source.
Conclusions
- It is important to consider the full product life cycle as well as the product performance when comparing the carbon footprint for different protein A resins.
- The resin manufacturing related carbon footprint per liter resin is smaller for the legacy resin. However, the carbon footprint across the entire protein A life cycle, when using 1 kg mAb as the reference flow, is smallest for MabSelect PrismA™ resin (Fig 9).
- High mAb capacity and long resin lifetime, of modern protein A resins improves the environmental sustainability of the protein A process.
- Alkaline stability of MabSelect SuRe™ and MabSelect PrismA™ resins enables the use of NaOH for CIP which reduces the carbon footprint compared to when using CIP including Gua-HCl for rProtein A Sepharose™ FF resin.
- The carbon footprint related to the cleanroom energy (HVAC) represents the hotspot in the mAb capture protein A step when using a US average electricity grid mix.
- The equipment footprint, cleanroom area, and consequently the energy for the cleanroom infrastructure is smallest for the MabSelect PrismA™ resin process.
Fig 9. Reduced carbon footprint per kg mAb with modern protein A resin processes.
CY43333
Data inventory protein A capture step (use phase)
Protein A cycle, buffers, and CIP
The buffers and volumes in the protein A capture step are presented in Table 3. We used buffers recommended for the Cytiva protein A resin platform (6). rProtein A is not alkaline stable so here we are assuming a three-step CIP procedure using 6 M Gua-HCl, detergent (0.5% Tween), and solvent (30% isopropanol) after each batch. MabSelect SuRe™ resin has good alkaline stability and can be cleaned with 0.1 M of NaOH. MabSelect PrismA™ resin has excellent alkaline stability and can be cleaned with 0.5 M of NaOH. The residence time (RT) for rProtein A Sepharose™ FF resin was assumed to be 8 min in all phases of the chromatography step. The RT for MabSelect SuRe™ and MabSelect PrismA™ resins was 4 min except for the CIP where the RT was 5 min.
After each batch the column was washed with 3 CV of storage solution (20% ethanol). Preparation of buffers and solutions included 10% extra volume.
Table 3. Buffers and volumes in column volume (CV) for the packed column in the Cytiva protein A platform process
| Phase | Buffer | Volume (CV) | Prepared volume, 10% extra (CV) | ||
| Equilibration and re-equilibration | 20 mM sodium (Na)-phosphate, 150 mM NaCl, pH 7.4 | 4.5 | 5.95 | ||
| Wash 1 | 20 mM Na-phosphate, 500 mM NaCl, pH 7.0 | 5 | 5.5 | ||
| Wash 2 | 50 mM acetate, pH 6.0 | 1 | 1.1 | ||
| Elution | 50 mM acetate pH 3.5 | 3 | 3.3 | ||
| Strip | 100 mM acetic acid, pH 2.9 | 2 | 2.2 | ||
| CIP | MabSelect SuRe™: 0.1 M NaOH MabSelect PrismA™: 0.5 M NaOH rProtein A Sepharose™ FF: 3-step CIP with detergent, solvent, and Gua-HCl |
3 | 3.3 | ||
| Sum | 18.5 | 20.35 | |||
DBC and assumed resin lifetime
The functional resin lifetime for MabSelect PrismA™ resin is assumed to be 200 cycles because it can be efficiently cleaned with high concentrations of NaOH without effect on capacity. The maximum lifetime for MabSelect SuRe™ resin is assumed to be 150 cycles since it is less alkaline stable compared to MabSelect PrismA™ resin and can be less efficiently cleaned. rProtein A Sepharose™ FF resin based on recombinant protein A is not alkaline stable and cannot be cleaned with high NaOH concentrations. Because of the lower ligand stability and less efficient CIP, the maximum resin lifetime for rProtein A Sepharose™ FF resin is assumed to be 100 cycles, that is, shorter than compared to MabSelect SuRe™ and MabSelect PrismA™ resins. For the high-demand process, it is assumed that the full lifetime of the resins is utilized.
Step yield
When calculating the amount of mAb that is produced/L resin, the yield over that process step needs to be taken into consideration. The aim of the mAb capture step is to remove the bulk impurities and concentrate the product at a high yield. In an affinity capture step, the target protein binds selectively to the ligand and can be eluted by decreasing the pH after washing off most impurities that do not have affinity to the protein A ligand. The yield in this process step can be very high, close to 100%. In this study, the yield was set to 97% for all resins.
Settled resin volume/L of the packed column
When calculating the amount of mAb that is produced/L of resin, the actual resin volume in a packed column needs to be considered. The bulk chromatography resin volume supplied to the customer in containers is defined by the gravity settled volume. The volume required to pack a column is determined by the compression factor (CF), which is defined as the bed height measured after settling by gravity divided by the packed bed height. For MabSelect SuRe™ and MabSelect PrismA™ resins in AxiChrom™ columns, the CF is 1.1, which means that 1.1 L of gravity-settled resin is required to pack a 1 L column. For rProtein A Sepharose™ FF resin, the CF is 1.15 so 1.15 L resin is required to pack a 1 L column (Table 4).
Table 4. Compression factors and resin volume per liter packed column
| Resin | Compression factor (CF) | Resin volume per liter of packed column (L) | |||
| MabSelect SuRe™ resin | 1.1 | 1.1 | |||
| MabSelect PrismA™ resin | 1.1 | 1.1 | |||
| rProtein A Sepharose™ FF resin | 1.15 | 1.15 | |||
Protein A cycling strategy and column sizes
In the high-demand mAb production scenario, the batch size going into the protein A capture step is assumed to be 50 kg, corresponding to a 10 000 L bioreactor at a titer of 5 g/L. Yield loss during the harvest was not accounted for here. The protein A process time was set to approximately 10 h.
The number of cycles per batch was obtained according to Equation 1.
Equation 1. Number of cycles per batch.
The column size was obtained according to Equation 2.
Equation 2. Column size.
The number of cycles per batch, and column sizes for the different protein A resins are presented in Table 5.
Table 5. High-demand (50 kg batch) cycling strategy, and column size
| Protein A resin | Cycle time (min) | Cycle time (h) | No. of cycles/batch | Protein A process time (h) | Process load (g/L) | Column size (L) | |
| rProtein A Sepharose™ FF resin | 169* | 2.81 | 3 | 9.6 | 28 | 595 | |
| MabSelect SuRe™ resin | 105 | 1.75 | 5 | 8.8 | 35 | 286 | |
| MabSelect PrismA™ resin | 152.7 | 2.55 | 3 | 7.6 | 60.6 | 275 | |
*With the three-step CIP procedure for rProtein A Sepharose™ FF resin, the cycle time is 2.8 h without CIP and 4.55 h with CIP and buffer washes in between CIP steps.
Equipment footprint
The aim of mapping of the equipment and to calculate equipment footprint is to determine the cleanroom area. The cleanroom area is assumed to be 10 times the equipment footprint, that is, the equipment footprint can occupy 10% of the cleanroom area.
It is assumed that buffers are prepared for each mAb batch. The buffer tank sizes were determined based on the volumes of each buffer. The product pool is assumed to be eluted in a volume of 1.5 CV for all three resins. Each eluate will be collected in separate tanks.
The buffer tank dimensions were obtained from the BioSolve software v9.0.8.16., which enabled calculation of the footprint of each buffer tank. The footprint of the tank for the cell culture supernatant was also included. The total tank footprint for the different protein A processes is presented in Table 6.
Table 6. Tank footprint
| Protein A process | Tank footprint (m2) | ||||
| rProtein A Sepharose™ FF resin | 31.5 | ||||
| MabSelect SuRe™ resin | 31.5 | ||||
| MabSelect PrismA™ resin | 22.0 | ||||
Chromatography systems
The flow rates in L/h determines the suitability of the chromatography system. The flow rates for the different protein A processes are calculated from the column volumes and the shortest residence times (RT), that is, highest flow rates in the protein A chromatography cycle. The flow rates are < 5000 L/h (Table 7), which allows the use of a BioProcess™ modular system 1½ inch (200 to 5000 L/h). The footprint of that system is approximately 3 m2.
Table 7. Protein A chromatography maximum flow rates
| Protein A resin | Column volume (L) | Shortest residence time (min) | Flow rate (L/min) | Flow rate (L/h) | |||
| rProtein A Sepharose™ FF resin | 595 | 8 | 74 | 4463 | |||
| MabSelect SuRe™ resin | 286 | 4 | 72 | 4290 | |||
| MabSelect PrismA™ resin | 275 | 4 | 69 | 4125 | |||
Chromatography columns
For the rProtein A Sepharose™ FF resin process, a 595 L column is needed to process a 50 kg mAb batch running three cycles in approximately 10 h. This requires an AxiChrom™ 2000 column. For MabSelect SuRe™ and MabSelect PrismA™ resin processes with column sizes of < 300 L, AxiChrom™ 1400 is the column of choice. The footprint of AxiChrom™ 2000 is approximately 6 m2 and the footprint of AxiChrom™ 1400 is approximately 4 m2.
Cleanroom (HVAC) energy consumption
The energy needed for the cleanroom infrastructure is dependent on the classification of the cleanroom. The cleanroom HVAC system consumes a lot of energy. Cleanroom classification grade C/ISO 7 is typical for the downstream purification steps (7). The energy for a grade C cleanroom in aseptic production is assumed to be 0.5 kW/m2 (8).
In this study we limited the facility/clean room to the downstream purification suite where the protein A capture step takes place. The cleanroom is maintained 24/7 throughout the year.
Energy consumption/yr = 0.5 (kW/m2) × 24 (h/d) × 365 (d/yr)/1000 (MWh/m2)
The energy for the cleanroom infrastructure per batch (or kg of mAb) is calculated from the size of the facility and the number of batches/yr. Here we assume that 94 batches/yr are produced.
Chromatography system energy consumption
The BioProcess™ 1 ½ inch modular system has an energy consumption of 3.5 kVA (VA=W). The protein A batch process time is assumed to be 10 h.
Cleaning of buffer tanks
Information about buffer tank cleaning was obtained from the BioSolve software v9.0.8.16. According to BioSolve, buffer tanks are only washed with water for injection (WFI).
Cleaning and sanitization of tanks for eluate pool
It is assumed that the product pool (eluate) is collected in a volume of 1.5 CV. Each eluate will be subjected to viral inactivation (VI) and therefore collected in separate tanks.
Buffer tanks were assumed to be cleaned only using water. However, the product pool tanks have been in contact with mAb, and therefore a more rigorous cleaning/sanitization is needed. The product pool tanks were washed with purified water (PW), caustic (0.2 M of NaOH), acidic rinse (0.4 M of phosphoric acid), and WFI (BioSolve).
Note that cleaning and sanitization of the tank for the cell culture supernatant was not included here.
Chromatography system
Environmental impact associated with cleaning and sanitization of the chromatography system was not included in this assessment due to lack of information of sanitization solution volumes for the large BioProcess™ modular systems. The same system is assumed for all three protein A processes and sanitization is assumed to be performed after each batch. Therefore, the impact from system sanitization will be the same for all three processes.
Related links
Designing in sustainability at Cytiva
MabSelect PrismA™ protein A resin
AxiChrom™ chromatography columns
BioProcess modular chromatography system
References
- The National Aeronautics and Space Administration (NASA). Accessed 16 April
- Decisions from Dubai: A review of COP28 outcomes - Climate Champions (unfccc.int)
- Health Systems taskforce | Sustainable Markets Initiative (sustainable-markets.org)
- 3 ways healthcare systems can reduce their carbon footprint, World Economic Forum, https://www.weforum.org/agenda/2022/11/3-ways-healthcare-systems-carbon-footprint/, Published 7 November 2022. Accessed 26 June 2023.
- Tad talks sustainability podcast. Episode 19: Life Cycle Assessment: The Key to Reducing Carbon Footprint — TadRadzinski
- How to use MabSelect PrismaA™ for antibody purification: a quick guide for developing a mAb capture step using this protein A resin. Get started with MabSelect™ PrismA protein A resin | Cytiva (cytivalifesciences.com)
- Jagschies G, Lindskog E, Łącki K, Galliher P. Biopharmaceutical Processing Development, Design, and Implementation of Manufacturing Processes, Günter Jagschies Eva Lindskog Karol Łącki Parrish Galliher. Elsevier; 2018.
- Cleanroom technology. Saving energy in cleanrooms (cleanroomtechnology.com)