How to Perform a First Pass Water System Sustainability Analysis

Stephen Hall and Sarah Kutz examine strategies to reduce the carbon footprint of water purification systems, including membrane-based water for injection generation and optimised flow rates, while exploring sustainability approaches and alternative designs

IN THEIR quest to reach sustainability goals, biopharmaceutical manufacturers are targeting their water systems. Water purification systems consume significant electrical and, often, thermal energy. Byproduct wastewater streams from water systems may contain high mineral content which may have a direct impact on the environment. These water systems also require constant monitoring and maintenance to ensure the water consistently meets stringent quality requirements.

There are numerous strategies to reduce the carbon footprint of both new and existing water purification systems. These range from the complex implementation of membrane-based systems instead of thermal water-for-injection (WFI) generation, to simpler measures like reducing circulation flow rates in water distribution loops. We explore the potential benefits – and risks – for various sustainability analysis approaches in both new construction and retrofits and take a look at some alternatives in system design through a case study.

 

More than carbon

A complete sustainability assessment has three dimensions and is called a “triple bottom line” (TBL) problem. The assessment is generally formulated as a multi-objective optimisation that considers the economic, environmental, and social aspects of the system’s demands. Many companies prioritise the economic dimension which includes capital and operating costs. Environmental concerns, usually assessed in terms of energy consumption or carbon emissions, are increasingly important. Social aspects such as the well-being of operators are often a secondary goal. The analysis can involve a linear programming method that juggles the competing goals among the dimensions; simplified approaches aggregate the impacts into a single composite metric.

Economic and environmental performance can be combined into an “eco-efficiency index” by measuring how efficiently a system uses resources to produce economic output. For example, if a process generates US$100,000 of economic value while emitting 10 t of CO2, the eco-efficiency would be US$10,000 per tonne of CO2.

Another way to simplify the problem is to score each dimension of the TBL using sustainability assessments with a qualitative comparison. For example, after calculating the carbon emissions from a process, the score might be 8–10 if there are minimal emissions, 4–6 for moderate impacts, or 1–3 for a dirty process. The scores for each dimension are then weighted according to the relative importance, based on input from stakeholders or from regulatory guidelines. This gives a total Life Cycle Sustainability Assessment (LCSA) score that can be used to compare competing approaches or systems.

Setting the scope

A well-defined scope is a critical component of sustainability analyses. Two scopes must be defined: temporal and spatial. Temporal scope comprises the life-cycle framework: does the analysis include the supply (cradle to gate), only the operations (gate to gate), or the disposal period (gate to grave)? Spatial scope is bounded by the physical extents of the system. For a water purification system, that could mean stipulating whether pre-treatment units are included in the system boundary, if installed spares should be considered, or if water users other than the pharmaceutical production equipment are a factor.

Establishing context

Energy calculations without context mean little. One way to provide perspective is to relate the consumption of water and its energy to the amount of product that the plant produces. This metric, known as carbon intensity (CI), works well for large facilities that continuously produce product. However, biopharmaceutical plants typically operate in batches, often manufacture sporadically due to product changeovers, and seldom provide a single production metric on which to base performance.

Another approach is based on the design capacity of the plant. The actual production will likely be lower than the design capacity, but the use of a sub-design production rate is too subjective to support a universally valid and justifiable metric. The difference is important in terms of sustainability, however: idle equipment still draws power. The water system consumes energy whether or not there is water demand – the calculated carbon intensity metric is higher when the plant is running below its design capacity.

Of course, energy can also be reported against the quantity of purified water that is produced. Published studies that compare membrane to thermal technologies for producing WFI often limit their horizon to a specific generation rate (eg kWh/L). This is a good starting point for a sustainability analysis, but a truly robust study – be it for design of a new system or for streamlining of an existing one – arises from careful consideration of some of the factors in the following sections.

Designing a new system

Water purification systems are typically sized to deliver a certain volume of water over a specified period. There is a trade-off between the generation rate and the size of the tank in the storage and distribution system.

For example, at peak production rates, engineers might calculate a need for 20 m³ of WFI over the course of a week. Additionally, they may determine that no more than 6 m³ would be consumed in a single day, with half of that amount drawn within a four-hour window. This is where the design trade-off factors in. If the tank has a working capacity of 4–6 m³, the generator could be sized to deliver approximately 250 litres per hour, or 6 m³ in 24 hours. However, if the tank is smaller, with a 1.5 m³ working volume, the generator would need to deliver 2 m³ in the four-hour window. This doubles the required delivery rate compared to the larger-tank scenario – increasing the energy consumption in the process.

The above is just one example of the intricacies of designing a water system through the lens of sustainability. To design water systems that have a minimal impact on the TBL, fundamental considerations include:

• right-sizing
• consumption by type of water
• temperature (eg effect of consumed water hot/ambient requirements, point-of-use vs subloop/dedicated tank methods to achieve temperature)
• plant production and future changes
• energy sources (eg electricity mix now, and in the future)
• technologies (eg membrane, thermal)
• parallel and N+1 installation
• sanitisation (eg heat or ozone)
• pumping energy (eg motor efficiency and hydraulic power)
• sampling and analysis
• energy recovery/wastewater recovery
• order-of-magnitude comparison (eg water system sustainability metrics compared to those of compressed air or heating, ventilation, and air conditioning for cleanrooms)
• contextualisation (eg holistic viewpoint of entire plant with its plant steam and other utilities)

Streamlining the system

Key to creating innovative sustainability measures in water systems – be they new or existing – is identifying waste. In an otherwise functional plant, ways of curtailing energy and water consumption are often not visible and may benefit from knowing where to look. Consider the following first:

• Use least purified water for different purposes (eg first rinse in cleaning)
• Optimise clean-in-place (CIP) flow and quantity
• Collect final rinse for use as next cleaning first rinse
• Assess the coefficient of performance (COP) of major equipment
• Maximise reverse osmosis (RO) recovery
• Avoid flushing to drain; monitor flow rate and duration before use or sampling
• Minimise losses during equipment standby
• Check source for humidification water
• Review pure steam usage
• Optimise sample coolers (both design and operation)
• Reclaim condensate
• Monitor total organic carbon (TOC) and conductivity sensor usage; minimise continuous flow to drain, return to system, or intermittent sampling
• Implement wastewater cooling strategies

Undertaking the analysis

Fortunately, engineers conducting water system sustainability analyses have access to numerous resources including research literature, institutional databases, and data from the automation historians for existing systems or similar plants. There are also various analytical tools available for sourcing and processing sustainability analysis calculations, such as OpenLCA software and the US Department of Energy MEASUR toolkit.

Once engineers have collected sufficient project information, quantifying sustainability impacts typically involves calculating commonly used metrics, as shown in Table 1.

Case study

This case study illustrates a sustainability analysis of biopharmaceutical water systems by comparing key sustainability metrics for three WFI generation technologies at a plant that requires both purified water and WFI. The metrics used for comparison are water consumption, wastewater generation, energy use, and carbon footprint. Multi-effect distillation and vapour compression are conventional thermal processes that have widespread industry acceptance. Membrane generation is an emerging technology that has been gaining popularity since its approval in the European Union in 2017.

We chose a gate-to-gate analysis. Research has shown that the energy consumption and carbon emissions associated with building water treatment plants and decommissioning them at the end of their lives are small regardless of the technology chosen compared to the operating costs. Therefore, simplifying the analysis to include only the operation and maintenance of the systems is a reasonable approach when comparing the relative sustainability impacts of alternative technologies.

Methodology

The study starts with quantification of the WFI and purified water demands, which are being used as a baseline. Peak flow rates, daily use, and annual totals are defined for the system. The system boundary extends from the delivery of drinking water through storage and distribution to the use points. Variations to the baseline are defined, showing how deviations from the baseline quantification might impact the results of the analysis. In order of increasing purity, water is needed for general cleaning, humidification, critical cleaning, blending into solutions that contact process equipment (eg phosphate buffers used in chromatography columns), and formulated into drug products (eg injectable drugs).

Factors that influence the sizing of equipment are varied in our model to provide a sensitivity analysis. The first step is to establish the baseline water requirements. Painting a picture of process water requirements in the facility starts with a list of points of use (POU) for purified water and WFI. These POU are then incorporated into the model using a Monte Carlo simulation that generates an hour-by-hour consumption schedule based on each POU’s duration of use, frequency per day, hot or ambient temperature, earliest and latest hours of use, and flow rate. Running this simulation for multiple days yields peak, daily average, and annual totals for the water. These quantities are cross-checked with plant data to perform a “reality test” on the results of the simulation.

Once the water demands have been calculated, equipment selections are made. This study presents a side-by-side comparison of multi-effect distillation, vapour compression, and membrane WFI generators based on vendor-supplied standard equipment.

Process description

A pharmaceutical fill-finishing plant produces oral suspensions in small bottles and sterile liquid formulations in IV bags. Both product families are formulated in stainless steel vessels that must be fully cleaned between batches. Batch sizes for bulk oral suspensions and IV solutions are 6,000 L and 2,000 L, respectively. The plant operates two shifts per day, five days per week, 48 weeks per year. The design production throughput is three batches per day for oral suspension and two batches per day for the sterile liquid.

WFI is required for formulating the sterile liquid, and for the final rinse when cleaning the sterile liquid tanks and tubing. All other water uses may be purified water grade.

Water consumption and cost

Using the water model methodology outlined in the preceding sections, the water consumption parameters are provided in Table 2. Configuring the model with each of the three WFI generation equipment options, the total water use can be found in Table 3. Wastewater totals can also be found in this table.

Figures 2a and 2b summarise the results of the analysis for the base case and the alternate design, respectively, showing the annual operating costs, energy consumption and carbon emissions. Costs are based on current typical values in the US, with plant steam generated from natural gas boilers; electrical steam generation doubles the cost and carbon emissions in the case study’s geographical location, where electricity has a higher carbon value than natural gas. The base case comprises separate storage and distribution systems for purified water and WFI.

Although 75% of the water used for product formulation could meet the quality standards of purified water, we developed an alternative design that eliminates the purified water storage and distribution system. Instead, all purified water users are assigned WFI. This change requires generating approximately four times more WFI than in the base case. However, by removing the reverse osmosis deionisation (RODI) unit (which is only necessary for multi-effect distillation), we can make an interesting comparison in terms of energy and cost.

While the multi-effect distillation solution still requires a reverse osmosis unit, making the alternate case less viable if multi-effect distillation is the preferred technology, we modified the reverse osmosis unit in our model. We downgraded it to a commercial standard, using plastic tanks and piping, which significantly reduces the initial cost. Although the operating costs are higher in the alternate case, the capital investment for vapour compression and membrane systems is lower due to the elimination of the RODI generator, the purified water storage tank, and distribution.

From this analysis, it is clear that vapour compression and membrane technologies have a distinct advantage over multi-effect distillation. However, a more detailed analysis is required before concluding that vapour compression or membrane is definitively the most advantageous option.

Discussion

Although the comparison between vapour compression and membrane may seem to be clearcut, there are nuances to consider. Vapour compression stills generally require less frequent maintenance and fewer operator (labour) hours than membrane. This study assumed that the WFI storage tank is maintained at ambient temperature which gives membrane an advantage since the thermal WFI units produce hot water. However, it is possible to recover some of the heat which would close the gap. If the project has a long time-scale, vapour compression offers an advantage due to its longer expected service life.

In terms of environmental impact, carbon emissions from energy utilisation comprise the most important element. The model used in this case study calculates the energy consumption involved in all three methods of WFI generation. We assume electrically powered membrane units, with thermal energy provided by gas-fuelled steam boilers for multi-effect distillation, and vapour compression equipment. It is important to note that the amount of carbon emissions associated with electric power depends on the regional energy “mix” (the combination of different primary energy sources such as coal, oil, natural gas, nuclear power, or renewable energy). In the US, where this test case was conducted, the Environmental Protection Agency (EPA) continuously monitors emissions from electricity generating units and publishes carbon equivalents used in this calculation.

A comprehensive economic analysis must account for various factors specific to the owner's policies and assumptions. These include the anticipated service life of the system, depreciation schedules, applicable taxes, labour absorption rates, and other overhead costs that impact operational expenditures. Additionally, considerations such as the cost of capital, energy efficiency improvements, and the potential for long-term savings or incentives play a critical role in evaluating the overall financial viability. Without integrating these elements, any cost comparison between alternative systems remains incomplete and may not accurately reflect the true economic impact on the organisation.

This is a first-pass assessment that provides direction for drilling deeper into the analysis of biopharmaceutical water system sustainability. It is not aiming to be comprehensive; instead, the goal is to bring to light areas of interest. For example, it may be surprising to note that in the studied region, the carbon equivalency rate for natural gas (0.185 kg CO2/kWh) is less than that of electricity (0.298 kg CO2/kWh) – resulting in an increased carbon footprint associated with electrification. Armed with that knowledge, further exploration of alternative energy sources for plant operation may be warranted. Another consideration is the maintenance component that varies from case to case and from one piece of equipment to another. An analysis such as this provides a starting point to coordinate with vendors, for instance, to explore the idea that microbial contamination requires increased monitoring in membrane-generated WFI.

Summary and conclusions

Meaningful sustainability analyses in the context of biopharmaceutical manufacturing encompass a multi-dimensional approach, considering economic, environmental, and social impacts within a well-defined scope. Engineers must balance these factors while addressing the unique challenges of biopharmaceutical production, such as batch operations and fluctuating production rates. Key strategies to enhance sustainability include right-sizing systems, optimising energy usage, and implementing innovative waste reduction measures.

While performing a sustainability analysis incurs no inherent risks, its findings must be subject to a rigorous exploration of potential risk before they can be implemented. Regulatory compliance comes first, as failure to meet stringent water quality standards can lead to costly out-of-specification investigations, product recalls, and potential market withdrawals. Water quality standards, especially regarding contamination, should be prioritised.

Ensuring manufacturing robustness should also be prioritised. For example, although implementation of parallel and N+1 systems may be a loss in terms of sustainability, it plays a large role in enhancing system reliability, minimising downtime, and reducing the risk of non-compliance. In other words, there needs to be a balance between the reduced carbon cost of minimising energy-intensive equipment and ensuring that equipment is readily available to meet production fluctuations.

What is clear is that sustainability initiatives in water systems offer significant potential benefits – it is identifying and instituting them that is the battle. Comprehensive analysis that balances sustainability goals with the practicalities of regulatory compliance and manufacturing robustness must leverage advanced tools and generate data-driven insights. This exercise comprises the heft of the design of water systems that minimise environmental impact while ensuring reliable and compliant production processes.

---

This article is provided for guidance alone. Expert engineering advice should be sought before application.