Why is mathematics important in water desalination? Because a reverse-osmosis plant must account for every stream, every dissolved mass, every unit of pressure and every kilowatt-hour. Seawater enters; fresh permeate and concentrated brine leave. Between those points, engineers use ratios, mass balances, membrane performance, pressure, energy and uncertainty to keep the process understandable and controllable.
For Singapore students, this is not a distant textbook problem. PUB describes desalinated water as Singapore’s fourth National Tap and states that the country uses reverse osmosis for desalination. Mathematics helps explain both the achievement and the trade-offs: how recovery raises freshwater output, why salt rejection is not the same as recovery, why pretreatment matters, and why producing more water does not remove constraints on energy, membranes or the marine environment.
The Short Answer: Desalination Is a Balance Sheet for Water and Salt
A desalination system has inputs, outputs and processes. The most basic model follows conservation of mass. If 100 cubic metres of feed enters during a period and storage inside the system is unchanged, the outlet streams must total 100 cubic metres. If 45 cubic metres becomes permeate, about 55 cubic metres remains as concentrate, ignoring smaller ancillary losses and streams for this simplified example.
The same accounting must be done for dissolved salt. Salt does not vanish because water crosses a membrane. Most salt is retained and leaves in the concentrate, while a small amount may pass into permeate. This is why the concentrate is saltier than the feed.
- Feed is water entering the membrane process.
- Permeate is the product stream that passes through the membrane.
- Concentrate or brine is the stream retaining most dissolved salts.
- Recovery is the fraction of feed water produced as permeate.
- Salt rejection describes how effectively the membrane prevents salt passage.
- Specific energy consumption is energy used per volume of product water.
These quantities sound similar until algebra separates them. Recovery compares water volumes. Rejection compares concentrations. Energy intensity compares energy with product volume. Good engineering decisions require all three.
Singapore’s Desalination Context
The PUB page on desalinated water identifies the opening of Singapore’s first desalination plant in September 2005 and says current reverse-osmosis desalination uses about 3.5 kWh of energy per cubic metre of drinking water. That figure is a dated operational context supplied by the authority, not a universal constant for every plant, operating condition or boundary definition.
PUB’s broader Singapore Water Story places desalination alongside water from local catchments, imported water and NEWater. Mathematics helps students understand why a diversified system is not a contest in which one source must “win”. Sources differ in rainfall dependence, energy needs, infrastructure, quality requirements, reliability and environmental effects.
When using a number such as 3.5 kWh/m³, always ask what is included. Does it refer to the reverse-osmosis stage, the full plant or another system boundary? Is it an average, target or design value? What date and conditions apply? The habit of reading boundaries is as important as the multiplication that follows.
Salinity and Concentration
Salinity is a measure of dissolved salts. It can be expressed in several ways, including grams per litre, milligrams per litre, parts per thousand or practical salinity measures, depending on context. These are not casually interchangeable; temperature, density and measurement conventions can matter. For classroom mass-balance work, grams of salt per litre of solution can provide a clear starting point.
Suppose a teaching model uses feed water containing 35 grams of dissolved salt per litre. One cubic metre equals 1,000 litres, so each cubic metre contains 35 × 1,000 = 35,000 grams, or 35 kilograms, of salt. If 10 m³ enters, the feed contains 350 kg of salt under the model.
The calculation shows why unit conversion must be explicit. Multiplying 35 g/L by 10 m³ without converting litres and cubic metres gives a meaningless hybrid. A unit factor, 1,000 L/m³, makes the structure visible and cancels correctly.
Concentration is not total amount
A small bottle can have high salinity but little total salt. A large tank can have lower salinity but more total salt. Total dissolved mass equals concentration multiplied by volume, provided the units and concentration definition are compatible. Engineers need both because membrane behaviour depends partly on concentration, while waste and recovery accounting depend on total mass flow.
Did You Know? When pure water is removed from a salty solution while most salt remains, the salt concentration rises even though no salt has been added. This is the same mathematical idea seen when a sauce reduces during cooking: the amount of water falls faster than the dissolved material.
Recovery Ratio: How Much Feed Becomes Product?
Recovery, commonly written R, is permeate flow divided by feed flow. If a system takes 200 m³/h of feed and produces 90 m³/h of permeate, recovery is 90 ÷ 200 = 0.45, or 45%. The concentrate flow in a simple steady water balance is 200 − 90 = 110 m³/h.
Recovery is not the same as efficiency in the everyday sense. A higher recovery produces more permeate from a given feed volume, but it also concentrates salts more strongly and can increase scaling risk or operating demands. There is no rule that “highest recovery is always best”. The feasible operating point depends on feed chemistry, pressure, membranes, pretreatment, energy, discharge conditions and reliability.
A reverse calculation
If a plant must deliver 72,000 m³/day at 45% recovery, idealised feed flow is 72,000 ÷ 0.45 = 160,000 m³/day. The idealised concentrate flow is 160,000 − 72,000 = 88,000 m³/day. Real planning may include plant availability, internal uses, flushing, pretreatment losses and reserve margins, so this is the core relationship rather than a complete design.
Percentage-point trap
If recovery rises from 40% to 45%, the increase is 5 percentage points. Relative to 40%, it is a 12.5% increase because 5 ÷ 40 = 0.125. Both statements can be true, but they answer different questions. Engineers, journalists and students should label which one they mean.
Salt Rejection and Salt Passage
Salt rejection compares feed and permeate concentrations. A common idealised formula is rejection = 1 − (permeate concentration ÷ feed concentration). If feed concentration is 35,000 mg/L and permeate concentration is 350 mg/L, the ratio is 0.01 and rejection is 99%.
Salt passage is the complementary percentage in this simple definition: 1%. This does not say that 1% of all feed salt necessarily appears in the permeate. Concentration ratio and total mass split are related but not identical because the permeate volume differs from the feed volume.
Check with a salt mass balance
Take 100 m³ of feed at 35 kg/m³. Total feed salt is 3,500 kg. At 45% recovery, permeate volume is 45 m³ and concentrate volume is 55 m³. If permeate concentration is 0.35 kg/m³, permeate salt is 45 × 0.35 = 15.75 kg. The remaining 3,484.25 kg is in the concentrate, giving an idealised concentrate concentration of 3,484.25 ÷ 55 ≈ 63.35 kg/m³.
The membrane rejection from concentrations is 1 − 0.35/35 = 0.99, or 99%. Yet the permeate carries only 15.75/3,500 = 0.45% of the feed salt because the permeate is both less salty and smaller in volume than the total feed. This worked example exposes a common misconception.
| Measure | Formula in this example | Result |
|---|---|---|
| Water recovery | 45 m³ ÷ 100 m³ | 45% |
| Concentration rejection | 1 − 0.35 ÷ 35 | 99% |
| Salt mass to permeate | 15.75 kg ÷ 3,500 kg | 0.45% |
| Concentrate salinity | 3,484.25 kg ÷ 55 m³ | 63.35 kg/m³ |
Osmosis, Pressure and Why Energy Is Needed
Osmosis describes the tendency of water to move across a suitable semipermeable membrane toward the more concentrated solution. Reverse osmosis applies pressure to drive water in the useful opposite direction while retaining most dissolved substances. The required pressure must overcome osmotic effects and provide enough driving force for flow through the membrane and system.
At introductory level, osmotic pressure for a dilute ideal solution is often represented by π = iCRT, where i represents an effective particle factor, C is molar concentration, R is the gas constant and T is absolute temperature. Real seawater is not a perfectly dilute ideal solution, so professional models use appropriate thermodynamics and empirical data. The school formula teaches dependence: higher dissolved concentration and temperature influence osmotic pressure.
Pressure multiplied by volume has units of energy. One pascal is one newton per square metre; multiplying by cubic metres gives newton-metres, or joules. This dimensional link explains why pumping pressurised water demands energy.
A scale estimate
If an idealised pump raises 1 m³ of water through a pressure difference of 5,000,000 Pa, the pressure-volume work is about 5,000,000 J, or 5 MJ, before efficiency and losses. Since 1 kWh equals 3.6 MJ, this is about 1.39 kWh of ideal pressure work. A real plant has additional stages and inefficiencies, while energy recovery devices can reclaim part of the concentrate’s pressure energy. The estimate is not a plant-performance claim; it is a dimensional bridge from pressure to electricity.
Specific Energy Consumption
Specific energy consumption, or SEC, is energy divided by product-water volume. If a plant uses 350,000 kWh in a period and produces 100,000 m³ of water, SEC is 3.5 kWh/m³. If daily product demand is 120,000 m³ at the same illustrative SEC, energy is 420,000 kWh per day.
This multiplication is useful but incomplete. Energy may vary with feed temperature, salinity, membrane condition, pressure, recovery and pretreatment. Plant-level boundaries also matter. A reported SEC should name what equipment and time period it covers.
PUB’s 3Rs decarbonisation strategy discusses reducing desalination’s energy requirement and describes research directions. It reports a current level around 3.5 kWh/m³ and a target direction toward 2.0 kWh/m³ through technologies under exploration. A target is not the same as a universal achieved value; dating and wording preserve the distinction.
Energy cost scenario
For an illustrative tariff of $0.18 per kWh, energy cost at 3.5 kWh/m³ is $0.63 per m³. At 2.8 kWh/m³ it is $0.504 per m³, a difference of $0.126 per m³. At 100,000 m³/day, that difference is $12,600 per day under the simplified assumptions.
The scenario is not a claim about an actual PUB electricity tariff or total water cost. It excludes capital, chemicals, labour, maintenance, distribution and environmental management. Its value is mathematical: it shows why a change that looks small per cubic metre can matter at system scale.
Energy Recovery Devices
The concentrate leaves the membrane still pressurised. Energy recovery devices transfer much of that pressure energy to incoming feed or another stream. If concentrate carries an idealised 1,000 units of recoverable pressure energy and a device transfers 950 units usefully, its transfer efficiency under that defined boundary is 95%.
But adding one device efficiency to another does not give overall plant efficiency. Pumps, motors, pipes, pretreatment and membranes interact. Multiplying stage efficiencies can be appropriate for a simple series: a motor at 96% and pump at 88% give 0.96 × 0.88 = 84.48% combined conversion across those two idealised stages. Whether that model is suitable depends on the energy flows being compared.
This develops systems thinking. A single high-performing component cannot erase losses elsewhere, and improving a component that contributes little to total consumption may have limited plant-level effect. Engineers often use a Sankey-style energy balance or detailed process model to find the larger terms.
Pretreatment, Fouling and Time
Reverse-osmosis membranes need suitable feed water. Pretreatment removes or controls particles, biological material and substances that might foul or scale membranes. Fouling can reduce permeability, increase pressure requirements and lead to cleaning or replacement.
Suppose a membrane train initially produces 100 units of flow at a standardised condition and later produces 92. The normalised decline is 8%. If raw flow fell while temperature also changed, comparing raw numbers might falsely attribute everything to fouling. Normalisation adjusts measurements to a common reference so trends are more comparable.
Exponential and linear models
A first classroom model might assume performance declines linearly by 0.5% of the initial value each month. After six months, it predicts a 3% decline. Another model might compound a 0.5% monthly retention factor: performance = 100 × 0.995⁶ ≈ 97.04. The answers are close at first but diverge over time.
Neither equation should be treated as a universal membrane law. Actual behaviour depends on operating conditions and interventions. The lesson is to identify whether a rate is applied to the original amount or the current amount, and then validate the chosen model against data.
Reliability and Capacity
Nameplate capacity is not the same as delivered annual production. A plant rated at 100,000 m³/day would produce 36.5 million m³/year if it ran at that output every day of a 365-day year. If planned and unplanned conditions lead to an average 90% capacity factor under a simplified definition, annual production is 32.85 million m³.
Water-security planning cannot rely on the average alone. Demand changes through the day and year. Plants need maintenance. Other sources vary. Storage and distribution impose constraints. Scenario models test combinations such as high demand, dry weather, maintenance outage and energy-price change.
This is a form of conditional reasoning: “If these assumptions hold, what follows?” A scenario is not a prediction that the conditions will occur. It is a structured way to see whether the system remains adequate if they do.
For a related treatment of rates and demand, read Why Mathematics? | Water Conservation, Flow Rates and Daily Use. Conservation and desalination address different parts of the water system, but both require volume, time and honest system boundaries.
Brine Dilution and Environmental Mathematics
Concentrate must be managed responsibly. Its salinity, temperature, residual treatment chemicals, discharge configuration, receiving environment and mixing conditions can matter. A dilution calculation is only the beginning of environmental assessment.
Consider an idealised mixture: 1 m³ of concentrate at 65 kg/m³ salt mixes perfectly with 9 m³ of seawater at 35 kg/m³. Total salt is 65 + 315 = 380 kg in 10 m³, giving 38 kg/m³. The model demonstrates mass balance. It does not prove that a real plume is instantly or uniformly mixed.
Real discharge analysis uses hydrodynamics, density differences, tides, currents, diffuser geometry and ecological evidence. Averages can hide near-field peaks. This is a valuable limit: correct arithmetic inside an unrealistic box can still produce a misleading environmental conclusion.
Students should also avoid the false claim that desalination is environmentally impact-free because the ocean is large. Scale, location, technology and regulation matter. Equally, it is unfair to declare every desalination project harmful without site-specific evidence. Mathematics helps keep evaluation proportional.
Worked Design Comparison
Imagine two fictional membrane configurations processing the same 100,000 m³/day feed.
Configuration A has 40% recovery, 99.5% salt rejection and SEC of 3.2 kWh/m³ product. It produces 40,000 m³/day permeate and 60,000 m³/day concentrate. Daily energy is 128,000 kWh.
Configuration B has 50% recovery, 99.2% rejection and SEC of 3.6 kWh/m³ product. It produces 50,000 m³/day permeate and 50,000 m³/day concentrate. Daily energy is 180,000 kWh.
If the requirement is 40,000 m³/day, A meets it with lower total energy in this simplified comparison. If the requirement is 50,000 m³/day and feed intake is capped at 100,000 m³/day, B meets the volume while A does not. Yet B also has lower rejection and higher SEC, so product quality, downstream treatment and energy must be checked.
| Criterion | Configuration A | Configuration B |
|---|---|---|
| Feed | 100,000 m³/day | 100,000 m³/day |
| Recovery | 40% | 50% |
| Product | 40,000 m³/day | 50,000 m³/day |
| Concentrate | 60,000 m³/day | 50,000 m³/day |
| Salt rejection | 99.5% | 99.2% |
| Specific energy | 3.2 kWh/m³ | 3.6 kWh/m³ |
| Daily energy | 128,000 kWh | 180,000 kWh |
There is no automatic winner. The correct decision uses constraints and priorities. This is optimisation: find a feasible option that performs well under several objectives, rather than maximising one attractive percentage.
Common Misconceptions
“Recovery and rejection are the same”
No. Recovery concerns how much feed water becomes permeate. Rejection concerns how much lower the permeate concentration is relative to feed concentration.
“A 99% rejection membrane removes 99% of total feed salt”
Not necessarily. That percentage is normally based on concentrations. The salt mass split also depends on stream volumes.
“Higher recovery is always more efficient”
No. Higher recovery can reduce intake and concentrate volume per product volume, but may raise concentration, pressure, scaling or treatment constraints. Evaluate the whole system.
“Desalinated water is pure H2O”
No. Product water is treated to meet applicable quality requirements and is conditioned for distribution. “Desalinated” does not mean chemically empty.
“Energy per cubic metre is the whole cost”
No. It omits capital, labour, chemicals, membranes, maintenance, intake, disposal and distribution unless the boundary explicitly includes them.
“A mass balance proves environmental safety”
No. It accounts for material. Environmental assessment also needs concentration patterns, exposure, ecology, timing and site-specific evidence.
Which Mathematics Matters?
Desalination connects school topics to an engineered system. Fractions become recovery. Ratio becomes concentration. Algebra becomes a mass balance. Graphs show membrane decline. Exponents describe compounding. Geometry and calculus can describe membrane area and flow. Statistics separates genuine performance change from measurement noise.
| School idea | Desalination mechanism | Transferable habit |
|---|---|---|
| Units | L, m³, mg/L, kg/m³, kWh | Convert before combining |
| Ratio | Recovery and rejection | Name numerator and denominator |
| Algebra | Feed = permeate + concentrate | Solve for missing streams |
| Percentage | Performance change | Separate points from percent change |
| Functions | Flux versus pressure or time | State the valid range |
| Statistics | Normalised trends and variation | Do not overreact to one value |
| Optimisation | Energy, quality, output, reliability | Respect multiple constraints |
| Geometry | Membrane area and module packing | Connect size to capacity |
This is why mathematics education matters beyond examinations. Students learn to conserve quantities, test plausibility and resist slogans. “More recovery”, “less energy” and “better quality” are incomplete until the denominator and boundary are clear.
A Safe Student Project
Build a spreadsheet model, not a homemade drinking-water system. Set cells for feed flow, recovery, feed salinity, rejection and SEC. Calculate permeate flow, concentrate flow, permeate concentration, salt masses and energy. Use invented parameters clearly labelled “illustrative”. Never drink water produced by an improvised membrane experiment.
Model equations
- Permeate flow = feed flow × recovery.
- Concentrate flow = feed flow − permeate flow.
- Permeate concentration = feed concentration × (1 − rejection).
- Feed salt mass rate = feed flow × feed concentration.
- Permeate salt mass rate = permeate flow × permeate concentration.
- Concentrate salt mass rate = feed salt mass rate − permeate salt mass rate.
- Energy rate = product flow × SEC.
Test recovery from 35% to 50% while holding other values fixed, then explain why the resulting table is a sensitivity analysis rather than a physical prediction. In a real plant, other variables would change too. Add a warning cell if mass-balance error exceeds a small tolerance; this teaches verification.
Next, vary rejection from 98.5% to 99.8%. Observe that a small percentage-point change can create a large relative change in salt passage. At 99% rejection, passage is 1%; at 99.5%, passage is 0.5%, which is half as large. This is a powerful percentage lesson.
Guidance for Parents and Students
Parents can turn news about water infrastructure into a numeracy conversation. Ask: What is the daily capacity? Is that feed or product? What recovery is assumed? Is the energy figure per cubic metre? Does the article state the year and system boundary? These questions develop practical judgement without requiring advanced chemistry.
Students should practise explaining a calculation in words. “I divided permeate by feed because recovery is the fraction of incoming water that becomes product.” That sentence is more valuable than a naked 45%. It shows the formula was selected for meaning.
When an answer looks surprising, perform three checks: confirm units, test conservation, and estimate a reasonable range. Permeate cannot exceed feed in a single-pass steady balance. Salt out should approximately equal salt in under the defined streams. A rejection percentage must sit between 0% and 100% under its ordinary definition.
For concentration reasoning in another context, Why Mathematics? | pH, Logarithms and Acid–Base Chemistry shows how a familiar chemical number can encode a non-linear scale. Desalination mostly uses direct concentration and balance, while pH reminds readers that not every chemical scale behaves linearly.
Careers and Learning Pathways
Water work can involve process engineering, environmental engineering, chemistry, microbiology, membrane research, mechanical and electrical engineering, instrumentation, operations, data science, policy, finance and communications. Some roles use advanced transport equations; others rely on dependable ratios, trends, controls and documentation.
Mathematics alone does not guarantee entry or employment. Students should check current course prerequisites and develop science, computing, writing, teamwork and safety awareness. A school spreadsheet project can reveal interest, but it does not lock a student into one career.
The strongest transferable skills are modest and durable: keep units, balance inputs and outputs, question boundaries, compare scenarios and explain uncertainty. Those habits are useful in almost every engineering system.
Frequently Asked Questions
What is reverse osmosis?
It is a membrane process in which applied pressure drives water through a semipermeable membrane while most dissolved salts are retained. Real plants include pretreatment, pumping, energy recovery, post-treatment and monitoring.
What does 45% recovery mean?
Under the stated boundary, 45 units of permeate are produced per 100 units of feed water. It does not mean 45% salt removal.
What does 99% salt rejection mean?
Under the usual concentration definition, permeate concentration is about 1% of feed concentration. Check the exact test conditions and definition.
Why is concentrate saltier?
Much of the water passes to the permeate while most salt remains. The retained salt is distributed through a smaller water volume.
Why not push recovery to nearly 100%?
As water is removed, concentration and osmotic pressure rise, and scaling, fouling, pressure and quality constraints become more demanding. Feasible recovery is technology- and feed-specific.
Is desalination independent of weather?
It is often described as climate-resilient because it does not depend directly on rainfall in the same way as catchment yield. It still depends on energy, infrastructure, seawater conditions, maintenance and environmental management.
Does desalination replace conservation?
No. Supply and demand measures work together. Conservation can reduce the volume that must be produced, treated and distributed.
Is a lower SEC always a fair comparison?
Only if system boundaries, feed salinity, recovery, product quality and operating conditions are comparable. Otherwise, the numbers may describe different tasks.
A Practical Learning Ladder
Begin with litres and cubic metres. Then learn concentration as mass per volume. Add steady water balances, followed by salt balances. Separate recovery from rejection. Connect pressure to energy through dimensional analysis. Finally, explore multi-objective optimisation and uncertainty.
- Foundation: Unit conversion and ratio.
- Balance: Feed, permeate and concentrate.
- Quality: Concentration, rejection and salt passage.
- Energy: Pressure, work and kWh/m³.
- Performance: Trends, normalisation and reliability.
- Systems: Cost, emissions, discharge and resilience.
- Judgement: Choose a solution under stated constraints.
At every stage, label illustrative values. Official figures should be linked to their dated source. A worked example should never be allowed to masquerade as a plant specification.
How a Control Room Uses the Mathematics
A working plant does not solve one equation each morning and then remain unchanged. Operators watch flows, pressures, conductivity, temperatures, tank levels and energy continuously. Each instrument produces a time series. Control limits, alarms and trends help people decide whether a change is ordinary variation, a sensor problem or a process condition requiring action.
Imagine that permeate conductivity rises slowly while feed salinity and temperature remain stable. One hypothesis is a change in membrane rejection; another is an instrument drift; another is a leaking seal. The team compares redundant measurements, calculates normalised performance and inspects which trains changed. A percentage trend is evidence, not diagnosis.
Alarm thresholds and hysteresis
If an alarm turns on at exactly 10.0 and off at exactly 10.0, a noisy signal near the boundary can switch repeatedly. Hysteresis uses separate thresholds—for example, alarm above 10.0 and clear below 9.5—to stabilise behaviour. The exact values require process knowledge; the mathematical idea is a state-dependent rule.
Rate-of-change alarms answer another question. A value can remain inside its ordinary range yet move unusually quickly. If pressure rises from 55 to 61 bar in three minutes, the average change is 2 bar/min. Whether that matters depends on equipment and procedure, but the calculation helps distinguish magnitude from speed.
Data reconciliation
Real meters rarely make feed exactly equal permeate plus concentrate because each has uncertainty. If feed reads 100.4 m³/h while outputs read 44.8 and 55.1, the apparent imbalance is 0.5 m³/h. Relative to feed, that is about 0.5%. Engineers decide whether the mismatch fits instrument uncertainty or suggests an unmeasured stream, timing difference or fault.
Blindly forcing one value to close the balance can hide a problem. Data reconciliation adjusts measurements using their uncertainties and conservation equations. More trusted instruments receive more weight. This is statistics serving physics: conservation provides structure, while uncertainty prevents fake exactness.
Scaling From a Membrane Coupon to a Plant
Research may begin with a small membrane coupon. Scaling to a pressure vessel and then a full plant is not simple multiplication. Flow per membrane area, concentration polarisation, channel hydraulics, recovery per element and fouling distribution change with geometry and staging.
If a coupon produces 20 litres per square metre per hour under a test condition, a naïve calculation for 10,000 m² gives 200,000 L/h, or 200 m³/h. That is a useful upper-level multiplication but not a design guarantee. A full system may operate at different pressure, temperature, recovery and feed quality; usable area and downtime matter.
A scale-up factor should therefore be validated. Engineers compare dimensionless groups or build pilot plants so the relevant transport processes remain comparable. When a student writes “multiply by area”, the next sentence should ask whether performance per area stays constant.
Staging and concentration
Membrane elements are often arranged so concentrate from one part feeds another. The later stage receives saltier water, so its osmotic pressure and flux conditions differ. Treating every element as identical to the first overstates production. A stage-by-stage spreadsheet can track flow and salt using the same balances repeatedly.
This is an excellent lesson in iteration. The output of one calculation becomes the input of the next. Small rounding errors can accumulate, so keep adequate precision internally and round only final reported values.
Ethical Communication of Water Numbers
Water is essential, so impressive numbers can easily become slogans. A plant capacity is not proof of daily delivery every day. A laboratory rejection percentage is not a guarantee for all contaminants. A low energy record at one condition is not a universal operating value. Responsible communication separates measured performance, design rating, model projection and aspiration.
When comparing technologies, disclose the functional unit. “Per cubic metre of permeate meeting the same quality” is fairer than “per cubic metre” when one system produces a different product. Disclose the date, feed conditions and boundary. Avoid describing one technology as sustainable merely because one metric improved.
Students can practise with a three-column note: fact, interpretation, scenario. An official capacity is a fact. Saying it strengthens supply diversity is an interpretation supported by system context. Calculating hypothetical energy at an invented tariff is a scenario. Keeping those columns separate makes writing more trustworthy.
Final Perspective: Conservation Before Optimisation
Desalination is an optimistic demonstration of human problem-solving. A membrane, pumps and carefully controlled treatment can turn seawater into a dependable source of drinking water. Yet the achievement remains grounded in constraints. Water and salt must balance. Pressure costs energy. Membranes change with time. Concentrate must be managed. Capacity must meet demand reliably.
Mathematics makes those constraints usable. It shows where material goes, how performance is defined and what trade-offs follow when one variable changes. It also prevents exaggerated claims: no percentage is meaningful without a denominator, and no efficiency number is fair without a boundary.
Read next: Why Mathematics? | Water Treatment, Chemical Dosing and Flow follows water after abstraction through another set of balances, rates and controls. Together, the articles show why mathematics is important in everyday life and national infrastructure: it turns a vital resource into quantities that can be checked, improved and stewarded responsibly.