Appendix A · Appendices

Notation, units, and dimensional checks

This appendix supplies a consistent starting basis for calculations that combine unit operation models. Several published correlations specify practical input units instead of explicitly dimensionless groups. Convert at the interface and preserve the original correlation basis in the parameter record. Rewriting the label on an axis affects displayed sizes and does not change a fitted coefficient's physical convention.

Stream and material quantities

SymbolMeaning in the core examplesCommon units
SDry solids mass flow ratet/h or kg/s
WWater mass flow ratet/h or kg/s
QTotal slurry volumetric flow ratem³/h or m³/s
MState inventory or accumulated amountt or kg
HᵢDry solids holdup of size class i in a millt
mᵢDry solids mass flow in class it/h
M[c,i]Dry mass rate of constituent c in class it/h
g꜀Bulk dry mass grade of constituent cfraction, %, ppm
ρₛ, ρwTrue particle density and water densitykg/m³ or t/m³
wₛ, φₛSolids concentration by mass and by volumefraction
F(d)Cumulative dry mass passing aperture dfraction
d(q)Size corresponding to passing fraction qµm or mm
F₈₀, P₈₀Feed and product 80% passing sizestypically µm
d₅₀, d₅₀cActual and corrected classifier cutstypically µm
pᵢ, p꜀ᵢPartition and corrected partition to a named productfraction
BᵢⱼMass proportion from parent j reporting to daughter ifraction
Sᵢ, DᵢSelection rate and discharge rate for size class itime⁻¹
τMean residence times, min, or h
kKinetic rate constantinverse time
PPower or partition function as indicated locallykW or fraction
εSpecific energykWh/t
vₛSettling velocity relative to the stated phase or framem/h or m/s
GSolids flux per equivalent horizontal tank areat/(m² h)

The symbol S is used for total solids rate and, with a subscript, for a size-specific selection frequency in a population balance. Under a defined time basis these quantities have different dimensions. The difference is a useful check when transcribing an equation. For example selection terms enter the balance as SᵢHᵢ in tonnes per hour if Sᵢ has units h⁻¹ and Hᵢ is in tonnes.

Percent content divided by 100 gives the fraction used in probability or balance equations. One ppm mass is 10⁻⁶ of the dry mass and equals 1 mg/kg or 1 g/t. For 65 t/h at 850 ppm, the tracer flow is 65 × 0.000850 = 0.05525 t/h, or 55.25 kg/h. If the assay is already expressed in mass per volume of slurry, multiplying it by dry solids flow changes its meaning. Replace the denominator with the matching volumetric flow.

Conversion factors that alter engineering predictions

ConvertOperation
mm to µmMultiply by 1,000
µm to mMultiply by 10⁻⁶
t/h to kg/sMultiply by 1,000/3,600
m³/h to m³/sDivide by 3,600
t/m³ to kg/m³Multiply by 1,000
kPa to PaMultiply by 1,000
mass fraction to mass percentageMultiply by 100
kWh/t to J/kgMultiply by 3,600
h⁻¹ to min⁻¹Divide by 60

Sieve aperture is a physical length. A mesh number without its standard and wire specification is not an exact length unit. For size benchmarks, enter or report the aperture instead of relying on the mesh label. Always preserve precision when moving between micrometres and millimetres. A round display can collapse two meaningful cut sizes into the same report label.

At the water properties used in many illustrative calculations, 1 m³/h corresponds numerically to 1 t/h. This equivalence changes if the liquid density differs from 1 t/m³. A brine stream of 75 m³/h at 1.09 t/m³ carries 81.75 t/h. If an operation withdraws 25 t/h water equivalent, the volume at that density is 22.94 m³/h. For the hydraulic calculation use slurry volume and mixture density, with enough physical detail to reflect the relevant rheology.

Geometry, pressure, and hydraulic work

The cross-sectional area of a circular pipe is πD²/4. With flow Q on a volume-per-second basis, superficial mean velocity is v = Q/A. At Q = 450 m³/h and diameter 250 mm, Q is 0.125 m³/s and A is 0.049087 m². Velocity is 2.546 m/s. Using 250 without converting the diameter would reduce the numerical velocity by six orders of magnitude.

Static head expressed as a pressure difference is Δp = ρgH. For a slurry density of 1,450 kg/m³ and head of 24 m, Δp is approximately 341 kPa using g = 9.80665 m/s². A water-based head conversion applied to this pressure would indicate about 34.8 m of water. Both results may appear on equipment documentation and report the same pressure on different reference fluid bases. Label the head basis explicitly.

Hydraulic power is QΔp or ρgQH using SI quantities. In the pipe example Q = 0.125 m³/s and 24 m slurry head give hydraulic power ≈42.66 kW. At a combined relevant efficiency of 0.68, shaft or electrical input on that efficiency's boundary is about 62.74 kW. An apparent head that includes friction and elevation already should not be added to the elevation a second time.

Time scale, hazard, and accumulation

An exponential kinetic law uses a dimensionless product kτ. With k = 0.30 min⁻¹ and τ = 5 min, the exponent is 1.5. In seconds these inputs become 0.005 s⁻¹ and 300 s, giving the same result. A kinetic test with recovery measured against elapsed time can estimate a rate only after any delay in the observation or collection system is considered.

A flow imbalance of 48 t/h maintained for 30 seconds accumulates 0.40 t. At 24 integration steps per minute the step is 2.5 seconds. Each explicit step would add 0.03333 t if the rate is constant. A process model that numerically multiplies 48 by 2.5 uses an hour-based rate with a second-based step and creates an artificial inventory. Apply the conversion once at the derivative or step interface. Testing a constant-flow case with an analytic answer helps locate this error before a closed dynamic run.

A hazard λ per operating hour has a probability interpretation over actual exposed time. For exponential exposure, no-failure probability over t is e⁻λᵗ. At λ = 0.020 h⁻¹ over 12 h of exposure, failure probability is 1 − e⁻⁰·²⁴ ≈ 21.34%. A line available for only 10 operational hours of a 12-hour window sees a different operating exposure. Calendar-based failure modes need their own clocks. A rate times duration approximates failure probability only for short intervals.

Competing size and rate definitions

A Rosin–Rammler scale a at 1 − e⁻¹ passing, screen aperture A, P₈₀ and corrected cyclone d₅₀c are all lengths. They are different model variables. The first is a CDF parameter. Aperture is geometry. P₈₀ is a property of a material stream. Corrected d₅₀c is an operating partition property. Unit consistency detects mistakes that cross dimensions and cannot detect mistakes that cross these definitions. A calculation record should attach population or equipment identity to every such value.

The mean volumetric residence time of a vessel is effective volume over volumetric flow on a consistent basis. The same ratio for a dry solid inventory and throughput gives a mean solids residence time in an appropriately mixed model. The two values can differ if processes selectively retain one phase. A thickener has rapid liquid turnover and possibly much slower solids transport. A simulation that forces a common mixed residence time can still match a steady solution while predicting inappropriate transient overflow quality.

Short dimensional audits

Cut scale. A sigmoid uses d/d₅₀c and a dimensionless slope. Changing both length values from µm to mm leaves every ratio unchanged. Changing only the feed values shifts modeled partition. In a converged recycle, the resulting error can be amplified by the return path.

Energy basis. At 210 t/h and 1,680 kW net grinding power, specific energy is 8 kWh/t. If recycled solids bring mill feed to 350 t/h at the same power, the specific energy per tonne of mill feed is 4.8 kWh/t. Both calculations divide the same power by different solids bases.

Moisture. A filter cake carries 19 t/h dry solids and 3 t/h water. Wet-basis moisture is 3/22 = 13.64%. Dry-basis moisture is 3/19 = 15.79%. Treating the former as the latter when estimating water return introduces a mass discrepancy.

Concentrate rate. A flotation model predicts 88% tracer recovery from a 120 t/h feed of 1.8% grade. At concentrate grade 18%, expected concentrate rate is (120 × 0.018 × 0.88)/0.18 = 10.56 t/h. Multiplying throughput by recovery would produce 105.6 t/h, treating mineral recovery as mass yield.