EMDS electricity calculator

Estimate installed motor power and electricity for agitated tanks, mixers, filters, dryers and rotary kilns, from a process you sketch below. Every number comes with a prediction interval from 799 vendor listings.

Preliminary version 2.0. Numbers may change before publication

Process flow diagram

Build the process

Everything is per batch. Each feed carries a mass, a dry-solids fraction and a bulk density; their volume is what the first units process. A unit takes in whatever it lists under “Receives from” and passes on what it makes: tanks and mixers pass everything on, filters keep the solids as a cake (the filtrate leaves the process), dryers remove moisture down to the outlet value, and kilns lose the mass fraction you set. Use a stream merge to combine streams and a stream split to divide one by fractions.

Per-kg results use the dry solids leaving the last units by default; for a liquid product, switch to total mass or enter your own product mass. Each unit has its own tier. Tier 1 needs no equipment size; Tiers 2–3 need Vm, the material held in the equipment, which for batch units defaults to one batch and for continuous units follows from the processing time and τ. Electricity depends only on what a unit processes; its outlet matters only downstream.

Set all units to

Feeds

Other electricity

Loads outside the units above, such as feed pumps, conveyors or HVAC. Always counted as electricity.

Add a unit:

Results

Medians with 90% ranges from 10,000 Monte Carlo draws. Units on the same model share its coefficient error; each unit adds its own scatter.

Where the energy goes

Each unit’s estimate, grouped by equipment type (or by electricity and heat), summed on the right. The Monte Carlo total in the cards above is higher, because the unit distributions are right-skewed.

Electricity by unit

Dots: each unit’s estimate. Whiskers: its exact 90% interval. Diamonds: entered heat. Log scale.

Distribution of the total

Histogram of the process total over 10,000 Monte Carlo draws, each the sum of one draw per unit. Darker bars fall between the 5th and 95th percentiles; the line marks the median.

Share by unit

Each unit’s share of the sum of unit estimates, electricity and heat when included.

What drives the uncertainty

Each unit’s contribution to the variance of the total, Cov(unit, total) / Var(total). Contributions add up to 100% and include errors shared by units on the same model.

Load factor sensitivity

Every unit set to the same LF, per batch. Unit electricity scales in proportion to LF; entered loads and heat do not.

Unit results

Sort by any column

Streams

Mass and volume per batch through each unit

Available equipment

Every model in the calculator, with its Tier 1 value, Tier 2 equation (default geometry) and valid range. The recommended model per equipment type follows Table 2 of the paper.

Method

Installed motor power intensity follows a power law of equipment capacity, fitted to outlier-free vendor listings (754 of 799). Electricity for each unit is

E = LF · (P/Vm) · Vb · τ / η(P)

  • Tier 1 needs no size. It resamples real units of the selected model and reports the geometric mean with the 5th–95th percentile range.
  • Tier 2 applies the fitted equation at your equipment size with the default geometry, LF = 0.60 and η from Seider et al. (Eq. 16.18) at rated power. Intervals are exact t-based 90% prediction intervals.
  • Motor efficiency at the unit’s total rated power P (in hp) follows Seider et al., Eq. 16.18: η = 0.80 + 0.0319 ln P − 0.00182 (ln P)². It is stated for 1–1,500 hp and extrapolated outside that range (Seider et al., 2016).1
  • Tier 3 lets you replace the default geometry, load factor and motor efficiency with your own values.
  • Vb is the volume processed: the inflow for tanks, mixers, dryers and kilns, and the cake volume for filters. For continuous units, Vm = (Vb / processing time) × τ.
  • Loads outside the vendor package, such as feed pumps, burners and heaters, are not included; each unit lists them.
  • Pooled filter and dryer models are not offered, because subtypes differ too much to pool.

1. Seider WD, Lewin DR, Seader JD, Widagdo S, Gani R, Ng KM. Product and Process Design Principles: Synthesis, Analysis, and Evaluation. 4th ed. Hoboken, NJ: Wiley; 2016. Eq. 16.18.

How to cite

If you use these estimates, please cite the paper and the calculator version:

This preliminary version follows the revised manuscript. Values may change before publication; the version number and Monte Carlo seed are stored in every download.