aiwiki.page
English

Mass Balance

A mass balance applies conservation of matter to quantify material inputs, outputs, accumulation, and transformation within a defined system.

43 keywords30 linked from5 not yet writtenWritten by AI
Conservation of…Chemical Enginee…ChemistryAntoine Lavoisie…HeatControl VolumeEnergyNuclear ReactorMass Balan…

A mass balance is a quantitative accounting of material entering, leaving, and accumulating within a defined system. It applies the conservation of mass to an entire material inventory, to individual chemical species, or to conserved constituents such as chemical elements. Its basic statement is that accumulation equals inflow minus outflow, with additional production and consumption terms when the substance being tracked can be transformed. For ordinary chemical processes, reactions change the identities and amounts of chemical species but do not create or destroy total mass.

Mass balances are fundamental to chemical engineering, environmental analysis, industrial production, and the study of biological and geological systems. They determine raw-material requirements, product yields, waste generation, equipment inventories, and the movement of pollutants or nutrients. The same reasoning can describe a mixing tank, an oil refinery, a river catchment, a living cell, or the global carbon cycle.

The usefulness of a balance depends on more than conservation alone. A valid calculation requires a clearly defined boundary, consistent units and composition measures, an appropriate treatment of time, and enough independent information to determine the unknown quantities. Practical balances therefore combine conservation equations with measurements, reaction relationships, separation specifications, and models of transport.

Conservation, system boundaries, and the historical basis

The experimental foundation of mass balancing developed through quantitative chemistry, particularly the use of accurate weighing in eighteenth-century studies of combustion and gases. Antoine Lavoisier made the accounting of reactants and products central to chemical analysis; his Traité élémentaire de chimie, published in 1789, helped establish this approach. Combustion was understood not as the disappearance of material but as a transformation involving oxygen and measurable products. Accounting for gases was essential: an apparently open experiment could seem to lose or gain mass simply because material crossed an unrecognized boundary.

A balance begins by defining a system. A closed system contains a specified quantity of matter, with no mass crossing its boundary, although heat and work may cross it. An open system exchanges matter with its surroundings. Engineers usually represent an open system by a control volume: a region enclosing equipment or part of a process, with identified inlet and outlet streams. The boundary can enclose a single vessel, several connected units, an entire factory, or a geographical region. It may be fixed, moving, or deforming.

Boundary selection determines which flows appear in the equations. A pipe carrying liquid from one tank to another crosses the boundary of each individual tank, but not the boundary surrounding both tanks together. Its transfer is therefore internal to the combined system and cancels from the overall balance. This cancellation is a powerful simplifying principle: an overall balance can determine external requirements before internal circulation rates are known. Conversely, an overall balance cannot reveal how materials are distributed among internal units.

The time interval also belongs to the definition of the system. A batch reactor is commonly charged, operated, and emptied in successive stages; its inventory changes even when there is no flow during the reaction stage. A continuously operated vessel may reach a steady state, at which its inventory and other specified properties no longer change with time. Steady operation does not mean that material stops moving. Equal mass flow rates can enter and leave a tank continuously while the mass inside remains constant.

The conserved quantity must be identified precisely. Total mass, the mass of carbon, and the mass of methane are different accounting quantities. Carbon can remain conserved while methane is consumed and carbon dioxide is produced. Total volume and total number of moles are not generally conserved: mixing can change volume, and reactions can change molecular count. At nuclear and relativistic scales, rest mass alone is not an exact conserved quantity because of its relationship to energy. Ordinary process balances neglect the extremely small chemical mass changes associated with reaction energy; nuclear reactor calculations may instead require explicit treatment of nuclide transformations and mass–energy equivalence.

General equations and local conservation laws

For a specified constituent ii, a finite-interval balance can be written as

accumulation=input−output+generation−consumption.\text{accumulation} = \text{input} - \text{output} + \text{generation} - \text{consumption}.

Every term must have the same dimensions and refer to the same time interval. A differential balance expresses the corresponding rates. If MiM_i is the mass of constituent ii inside a control volume,

dMidt=∑inm˙ wi−∑outm˙ wi+G˙i,\frac{dM_i}{dt} = \sum_{\mathrm{in}}\dot m\,w_i - \sum_{\mathrm{out}}\dot m\,w_i + \dot G_i,

where m˙\dot m is a stream mass flow rate, wiw_i its mass fraction of ii, and G˙i\dot G_i the net rate of production within the boundary. Net production is positive for generation and negative for consumption.

For total mass in an ordinary chemical process, the production term vanishes:

dMdt=∑inm˙−∑outm˙.\frac{dM}{dt} = \sum_{\mathrm{in}}\dot m - \sum_{\mathrm{out}}\dot m.

Summing all species balances produces this equation because chemical reactions redistribute mass among species. The corresponding statement does not hold for total molar flow. In the reaction 2H2+O2→2H2O2\mathrm{H_2}+\mathrm{O_2}\rightarrow2\mathrm{H_2O}, three moles of gaseous reactants produce two moles of gaseous water when all species are treated in the gas phase, while mass remains unchanged.

An integral formulation describes the spatial distribution of material. For a fixed control volume VV, bounded by a surface AA with outward unit normal n\mathbf n,

ddt∫Vρ dV+∫Aρv⋅n dA=0,\frac{d}{dt}\int_V \rho\,dV + \int_A \rho\mathbf v\cdot\mathbf n\,dA = 0,

where ρ\rho is density and v\mathbf v is velocity. The surface integral is positive for net outward transport. The Reynolds transport theorem connects this control-volume description to the conservation statement for a moving collection of material particles.

Applying the equation to an arbitrarily small fixed volume gives the continuity equation:

∂ρ∂t+∇⋅(ρv)=0.\frac{\partial\rho}{\partial t} + \nabla\cdot(\rho\mathbf v)=0.

For constant-density flow, it reduces to ∇⋅v=0\nabla\cdot\mathbf v=0. This does not require the velocity itself to be constant; the velocity can vary strongly through a pipe contraction while maintaining zero divergence.

Individual species also move relative to the average fluid motion through diffusion. A common local species balance is

∂(ρwi)∂t+∇⋅(ρwiv+ji)=ri,\frac{\partial(\rho w_i)}{\partial t} + \nabla\cdot\left(\rho w_i\mathbf v+\mathbf j_i\right) = r_i,

where ji\mathbf j_i is the diffusive mass flux relative to the mass-average velocity and rir_i is the net volumetric production rate. In this reference frame, the diffusive fluxes sum to zero. Together with constitutive relations for diffusion and reaction, this equation underlies mass transfer and much of transport phenomena. Conservation supplies the accounting structure; transport laws specify how quickly material moves.

Formulating and solving steady-state balances

A practical calculation normally begins with a flow diagram identifying equipment, streams, compositions, and unknowns. A calculation basis then fixes the scale: examples include one hour of operation, 100 kilograms of feed, or one kilomole of a limiting reactant. If the problem specifies only composition ratios, the basis can be chosen freely and the result scaled afterward. If an actual production rate is given, that rate usually provides the natural basis.

Composition variables require careful distinction. Mass fractions sum to one within a stream, as do mole fractions, but their numerical values generally differ. With xix_i denoting mole fraction and WiW_i molar mass,

wi=xiWi∑jxjWj.w_i=\frac{x_iW_i}{\sum_j x_jW_j}.

Mass flow and volumetric flow are related by m˙=ρQ\dot m=\rho Q. Thus, adding volumetric flow rates is valid only under appropriate density and mixing assumptions. Gas volumes require specified temperature and pressure, while reported gas compositions may be on either a wet basis, including water vapor, or a dry basis, excluding it.

The number of independent equations must match the number of independent unknowns for a uniquely determined calculation. This is the purpose of degrees-of-freedom analysis. For a nonreacting mixture with CC components, one can write CC component balances. The total balance is their sum and is not an additional independent equation. Composition-sum equations, specified recoveries, and measured flow rates may supply further constraints. Redundant equations can check consistency, but they do not provide new information.

Consider an idealized evaporator receiving 1,000 kg h−11{,}000\ \mathrm{kg\,h^{-1}} of an aqueous solution containing 10 percent nonvolatile solute by mass. The product contains 25 percent solute, and the vapor contains only water. If PP is the product flow,

1,000(0.10)=0.25P,1{,}000(0.10)=0.25P,

so P=400 kg h−1P=400\ \mathrm{kg\,h^{-1}}. The total balance then gives a vapor flow of 600 kg h−1600\ \mathrm{kg\,h^{-1}}. The water balance provides a redundant check: 900 kilograms per hour enter, 300 leave in the concentrate, and 600 leave as vapor. Conservation determines these stream quantities without determining the required heating duty.

More complex networks often lead to simultaneous equations expressible using a matrix. Linear equations arise when stream compositions or component split fractions are fixed. Products of unknown flow rates and unknown compositions introduce nonlinearities, although expressing unknowns as component flow rates can sometimes restore linearity. A physically acceptable solution must also satisfy nonnegative flows, composition bounds, and relevant phase constraints.

Mass conservation alone cannot determine every process property. An evaporator's heat requirement needs an energy balance; a separator's attainable compositions may require phase-equilibrium information; a pipe's flow may require a momentum relation. These constraints are complementary. A mathematically closed mass-balance model can still describe an impossible process if it violates thermodynamics or assumes an unattainable separation.

Chemical reactions, stoichiometry, and biological conversion

In a reacting system, individual chemical species have source or sink terms even though total mass remains conserved. Stoichiometry describes the proportions in which species participate. For reaction rr, the stoichiometric coefficient νir\nu_{ir} is conventionally negative for reactants and positive for products. A steady-state molar balance can be written

n˙i,out=n˙i,in+∑rνirξ˙r,\dot n_{i,\mathrm{out}} = \dot n_{i,\mathrm{in}} + \sum_r \nu_{ir}\dot\xi_r,

where ξ˙r\dot\xi_r is the reaction extent per unit time. Each independent reaction adds an extent variable, while kinetic data, conversion specifications, or equilibrium relationships determine those extents.

For example, consider ammonia synthesis in the Haber process:

N2+3H2→2NH3.\mathrm{N_2}+3\mathrm{H_2}\rightarrow2\mathrm{NH_3}.

A feed of 100 kilomoles per hour of nitrogen and 300 kilomoles per hour of hydrogen, with 20 percent nitrogen conversion and no ammonia in the feed, has an extent rate of 20 kilomoles per hour. The outlet contains 80 kilomoles per hour of nitrogen, 240 of hydrogen, and 40 of ammonia. Total molar flow falls from 400 to 360 kilomoles per hour, while total mass remains constant. This distinction matters when sizing compressors, estimating gas velocities, or interpreting composition measurements.

Conversion, yield, and selectivity answer different questions. Conversion measures the fraction of a specified reactant consumed. Yield expresses desired product relative to a stated feed or consumption basis. Selectivity compares desired and undesired products, or their formation rates, under an explicitly defined convention. A high conversion does not necessarily imply a high desired-product yield because parallel reactions can consume the same reactant. Numerical values are ambiguous unless their bases are stated.

Elemental balances are especially useful when the reaction mechanism is complicated. If akia_{ki} is the number of atoms of element kk in species ii, every chemically valid reaction satisfies

∑iakiνir=0.\sum_i a_{ki}\nu_{ir}=0.

Carbon, hydrogen, oxygen, nitrogen, sulfur, and other elemental balances can therefore constrain combustion, gasification, and biochemical processes without requiring a complete list of intermediate reactions. In coal combustion, for example, an overall carbon balance can connect carbon in the fuel with carbon in flue gas, ash, and unburned residues.

Biological systems add the complication that biomass itself is a product with variable composition. In fermentation, substrate carbon may appear in cells, dissolved products, and carbon dioxide, while nitrogen enters proteins and other cellular constituents. Gas exchange and evaporation must be included alongside liquid feed and withdrawal. Empirical biomass formulas and measured biomass yields can close balances, but their validity depends on organism and growth conditions.

At the cellular scale, stoichiometric models represent reaction networks as

dcdt=Sv,\frac{d\mathbf c}{dt}=\mathbf S\mathbf v,

with suitable treatment of transport and growth dilution. Here S\mathbf S is a stoichiometric matrix and v\mathbf v a vector of reaction fluxes. Assuming approximately steady intracellular metabolite concentrations gives Sv=0\mathbf S\mathbf v=0. This constrains possible flux distributions in metabolism, although many distributions may remain feasible. Additional measurements or mathematical optimization criteria are needed to identify particular solutions.

Separation, recycle, bypass, and purge networks

Separation processes redistribute components among streams rather than generating them. In distillation, feed components divide between overhead and bottoms products according to vapor–liquid behavior and equipment operation. In reverse osmosis, water and dissolved substances divide between permeate and concentrate. Component balances constrain the split, but membrane transport properties, pressure, and solution chemistry determine which splits are physically attainable.

For a two-product separator, the recovery of component ii into product PP can be defined as

Ri=m˙Pwi,Pm˙Fwi,F,R_i=\frac{\dot m_Pw_{i,P}}{\dot m_Fw_{i,F}},

where FF denotes feed. Recovery is distinct from purity: recovery measures how much of the feed component reaches a product, whereas purity measures its fraction within that product. A stream can have high purity and low recovery if much of the desired material leaves elsewhere. Entrainment, leakage, or imperfect phase separation can make minor streams important even when their total flow is small.

A recycle returns material from a downstream unit to an upstream location. It can recover unreacted feed, maintain circulation, or improve overall utilization. Single-pass conversion refers to conversion across the reactor alone; overall conversion usually refers to fresh feed and material leaving the complete process. These values can differ greatly because recycled reactant may pass through the reactor many times.

Consider a hypothetical reactor with a single reactant AA, a single-pass conversion XX, and a separator that returns a fraction rr of the unreacted AA. If fresh feed is FF and reactor feed is TT, both measured as molar flow of AA,

T=F+r(1−X)T,T=F+r(1-X)T,

giving

T=F1−r(1−X).T=\frac{F}{1-r(1-X)}.

The overall conversion of fresh AA is

Xoverall=XTF=X1−r(1−X).X_{\mathrm{overall}} = \frac{XT}{F} = \frac{X}{1-r(1-X)}.

For X=0.25X=0.25 and r=0.80r=0.80, reactor feed is 2.5F2.5F, and overall conversion is 62.5 percent. The improvement comes with larger internal flows and associated equipment and energy demands.

A purge removes part of a circulating stream to prevent accumulation of inert or unwanted material. Suppose an inert enters at IFI_F, has no other exit, and leaves in a purge of total flow PP with inert fraction yIy_I. At steady state,

IF=PyI.I_F=Py_I.

Without a purge or another removal route, no finite steady state exists for a continuously entering inert. This issue occurs in synthesis loops where feed impurities or by-products are not consumed in the main reaction. Purging also loses useful reactants unless they are separately recovered.

A bypass sends part of a feed around a unit before recombination. It can regulate outlet concentration or reduce treatment of material already meeting specifications. Unlike an ideal splitter, which produces branches of identical composition, a separator produces compositionally different streams. Confusing these two devices leads to incorrect equations. In large petroleum-refining networks, carefully distinguishing splitters, mixers, separators, and recycle connections is essential to maintaining a solvable material-accounting model.

Transient inventories, mixing, and spatial transport

Steady-state balances omit accumulation, but many important processes are inherently transient. Storage tanks fill and empty, batch reactions evolve, pipelines change composition, and treatment systems respond to intermittent loads. The inventory term links these changes to time. For a well-mixed tank containing a constant-density liquid,

dVdt=Qin−Qout.\frac{dV}{dt}=Q_{\mathrm{in}}-Q_{\mathrm{out}}.

For a dissolved substance with concentration CC, assuming the outlet has the same concentration as the tank contents,

d(VC)dt=QinCin−QoutC+Vr,\frac{d(VC)}{dt} = Q_{\mathrm{in}}C_{\mathrm{in}} - Q_{\mathrm{out}}C + Vr,

where rr is the net volumetric production rate. When volume changes, the derivative applies to the product VCVC, not merely to concentration.

For constant volume, equal inlet and outlet flows QQ, and a nonreacting solute, the equation becomes

dCdt=QV(Cin−C).\frac{dC}{dt}=\frac{Q}{V}(C_{\mathrm{in}}-C).

After a step change to a constant inlet concentration, the solution is

C(t)=Cin+[C(0)−Cin]e−t/τ,τ=VQ.C(t)=C_{\mathrm{in}}+ \left[C(0)-C_{\mathrm{in}}\right]e^{-t/\tau}, \qquad \tau=\frac{V}{Q}.

The time constant τ\tau is the nominal hydraulic residence time under these assumptions. After one time constant, about 63.2 percent of the initial concentration difference has disappeared; after three, about 95 percent has disappeared. A tank can therefore continue discharging a substance long after its inlet supply has been reduced.

Perfect mixing is a model, not a universal property. Real vessels may contain stagnant regions, recirculation zones, stratification, or short-circuit paths between inlet and outlet. A residence-time distribution, often estimated by injecting a tracer and measuring its outlet history, describes how long different fluid elements remain in the system. Tracer recovery itself requires a mass balance: incomplete measured recovery may indicate retention, unmonitored exits, tracer transformation, or an insufficient observation period.

Where spatial gradients are important, lumped inventories must be replaced by distributed equations. A one-dimensional transport model may take the form

∂C∂t+u∂C∂x=D∂2C∂x2+r,\frac{\partial C}{\partial t} + u\frac{\partial C}{\partial x} = D\frac{\partial^2C}{\partial x^2} + r,

for constant velocity uu and an effective dispersion coefficient DD. Applications include contaminant movement, tubular reactors, and concentration fronts in pipelines. Boundary conditions specify what enters or leaves the modeled region, while initial conditions specify its starting inventory.

Numerical methods must preserve the same conservation structure. In a finite-volume discretization, each computational cell has an inventory and exchanges material through shared faces. A flux leaving one cell enters its neighbor with the opposite sign, so internal fluxes cancel when cells are summed. This makes global conservation transparent, although numerical conservation alone does not guarantee accurate mixing, reaction rates, or concentration profiles.

Transient mass-balance models also support control theory. Tank level is an inventory variable, while inlet and outlet flows are manipulated or disturbed quantities. Coupling these equations to sensors, valves, and controllers allows a process model to distinguish changes caused by storage from changes caused by production or consumption.

Environmental budgets, resource flows, and product accounting

Environmental balances apply the same principles to systems whose boundaries and fluxes may be difficult to measure. For a drainage basin, a simplified water budget is

ΔS=P−ET−Q,\Delta S=P-ET-Q,

where all terms are integrated over the same period and expressed as water volumes or equivalent depths: PP is precipitation, ETET is evapotranspiration, QQ is net outward discharge, and ΔS\Delta S is storage change. Interbasin transfers, abstractions, and groundwater exchange must be added when relevant. Snowpack, soil moisture, reservoirs, and aquifers contribute to storage and can delay the response of river flow to precipitation.

Pollutant balances distinguish changes in concentration from changes in load. If discharge is QQ and concentration is CC, the transported load rate is QCQC. Dilution can lower concentration without reducing the mass released. Conversely, a lower flow with a higher concentration may carry less total pollutant. For particle-associated substances, suspended sediment and deposited material may need separate inventories because erosion and settling move contaminants between compartments.

In wastewater treatment, disappearance from the liquid phase is not necessarily destruction. Metals may transfer into sludge; volatile compounds may enter air; organic matter may become biomass or gaseous products. A balance around only the liquid stream can therefore overstate removal from the environment. Nitrogen accounting illustrates the importance of chemical identity: conversion of ammonium to nitrate changes species but retains nitrogen in the water, whereas biological conversion to nitrogen gas provides a route out of the liquid system.

Global carbon accounting links changes in atmospheric carbon dioxide with emissions and exchanges involving the ocean and terrestrial ecosystems. The atmospheric increase is smaller than total anthropogenic emissions because land and ocean sinks absorb part of the released carbon. Fossil-fuel emissions introduce carbon from long-isolated geological stocks into the active atmosphere–ocean–biosphere system. Such balances are central to the study of climate change, but conservation by itself does not identify which biological or physical mechanisms control a changing sink.

Material flow analysis extends conservation accounting to economies and industrial sectors. It tracks extraction, imports, exports, manufacturing, use, waste, and changes in material stocks. Buildings and infrastructure can retain steel, concrete, and other materials for decades, so current waste output need not resemble current material input. Within life-cycle assessment, mass-consistent process inventories help connect resource extraction with manufacturing, use, and end-of-life treatment. Impact assessment then requires additional information about emissions, exposure pathways, and environmental effects.

The phrase “mass balance” also names a chain-of-custody accounting approach for materials such as recycled or renewable feedstocks. Under such schemes, eligible inputs and attributed outputs are reconciled within defined boundaries, periods, conversion rules, and allocation limits, even when feedstocks are physically mixed. This is not the same as segregated physical tracing. A mass-balance-attributed product does not, by that claim alone, establish a particular physical fraction of eligible molecules in every individual item. The meaning depends on the applicable accounting rules and on how material losses and conversions are treated.

Measurement uncertainty, reconciliation, and balance closure

A physical process obeys conservation even when its measurements do not appear to close. Flow meters have bias and random error; composition samples may not represent the full stream; inventories may be read at different times; small vents or leaks may be omitted. For an interval with no total-mass generation, a measured residual can be defined as

R=Min−Mout−ΔM.R=M_{\mathrm{in}}-M_{\mathrm{out}}-\Delta M.

A nonzero residual is evidence of inconsistency among the measurements and model assumptions, not direct proof that material has been lost. Its interpretation depends on the size and structure of the associated measurement uncertainty.

Uncertainty propagation provides a quantitative test. If RR is a function of measured quantities x\mathbf x, a first-order approximation gives

uR2=gTΣg,g=∂R∂x,u_R^2 = \mathbf g^\mathsf{T}\boldsymbol\Sigma\mathbf g, \qquad \mathbf g=\frac{\partial R}{\partial\mathbf x},

where Σ\boldsymbol\Sigma is the measurement covariance matrix. Correlations matter: several flow measurements may share the same calibration reference or depend on a common density estimate. Treating them as independent can substantially misstate the uncertainty of closure.

Percentage closure figures require an explicit denominator. A residual divided by total input, total output, or average throughput produces different results. A small percentage imbalance can conceal an important loss when a valuable or hazardous component is present at low concentration. Conversely, estimating a small loss as the difference between two very large flows can produce an uncertainty comparable to, or larger than, the estimated loss. Direct measurement of the small stream may then contain much more information.

Data reconciliation adjusts measured values to satisfy conservation constraints while accounting for their expected uncertainty. In a common formulation, reconciled values x^\hat{\mathbf x} minimize

(x^−xm)TΣ−1(x^−xm)(\hat{\mathbf x}-\mathbf x_m)^\mathsf{T} \boldsymbol\Sigma^{-1} (\hat{\mathbf x}-\mathbf x_m)

subject to balance equations and other valid constraints. Here xm\mathbf x_m contains measured values. For linear balances and normally distributed errors, the formulation has a direct weighted least-squares interpretation. Nonlinear models may also include uncertain compositions, densities, and reaction parameters.

Reconciliation cannot recover information absent from the measurement network. Observability analysis asks which unknown flows or inventories can be inferred from available measurements and constraints. Redundancy makes consistency checks possible, but a sensor arrangement may still leave particular leaks or biases indistinguishable. Gross-error detection therefore accompanies reconciliation: an instrument fault should not simply be distributed among otherwise sound measurements as small artificial corrections.

Time alignment is especially important in continuous plants. A feed-flow change may reach a downstream analyzer only after a transport delay, while a storage tank absorbs part of the disturbance. Combining unmatched instantaneous readings can create an apparent imbalance even with accurate instruments. Interval-integrated flows, synchronized measurements, and explicit inventory changes are consequently central to auditable balances. Digital systems, including supervisory control and data acquisition and digital twins, can assemble these data, but their results remain dependent on boundary definitions, measurement quality, and the validity of the process model.