Chemical & Process Engineering

Highly specialised requests are first checked for feasibility from this information. Please do not send confidential data through the public form; describe it first and agree secure transfer afterwards.

1. 01Enquiry and requirement assessment
2. 02Technical scoping and proposal
3. 03Execution
4. 04Quality review and delivery, with an agreed round of revisions

Read the full project workflow.

In depth: how the work is done

1. Mass and energy balance modelling

   Balances provide the backbone of process analysis. Steady-state balances size flows and duties, while dynamic balances describe start-up, disturbances and control. Consistency checks, such as closing element balances on measured data, reveal measurement errors. Data reconciliation adjusts measurements to satisfy balances, using their uncertainty as weights.

2. Kinetic parameter estimation

   Rate laws and activation energies are estimated by fitting model predictions to concentration-time data by nonlinear regression. Parameters are often correlated, such as pre-exponential factors and activation energies, so reparameterisation improves estimation. Confidence regions, residual diagnostics and model discrimination between candidate mechanisms indicate whether the data support a particular model. Extrapolation beyond the tested temperature or concentration range is flagged.

3. Reactor and separation modelling

   Ideal reactor models (batch, CSTR, plug-flow) are starting points; real reactors have non-ideal flow, which residence-time distributions characterise. Separation units such as distillation, absorption and membranes are modelled with equilibrium and rate-based approaches. Thermodynamic property models strongly influence results, so their selection and validity range are examined.

4. Process optimisation and design of experiments

   Experimental optimisation uses factorial and response-surface methods to find favourable temperature, pressure, concentration and time. Model-based optimisation uses mathematical programming on process models, with constraints on safety and quality. Both approaches depend on the correctness of models, and we recommend confirmation runs and sensitivity analysis before claims are made.

5. Scale-up and sustainability considerations

   Scale-up changes heat and mass transfer limits, mixing times and safety margins. Dimensionless numbers such as Reynolds, Damköhler and Péclet help identify controlling mechanisms. Sustainability analysis considers energy use, solvent choice, waste and life-cycle impacts, with system boundaries stated. Laboratory results are presented as such, with scale-up questions listed explicitly.

6. Laboratory data management for process studies

   Process experiments generate time series from many sensors plus offline analyses. Keeping a run log that records set-points, feed batches, calibration of analysers and any disturbances makes later interpretation reliable. Time alignment between online and offline measurements should be documented, and analytical methods should have known detection limits and repeatability. Replicate runs, including at least one centre-point repeat, estimate pure error and support lack-of-fit testing. These habits protect against the common problem of fitting models to data whose conditions were not recorded.

7. Choosing between experiment and model-based effort

   Process research can spend its budget on more experiments or on better models, and the best balance depends on how much is already known. When mechanisms are well understood, a model can screen conditions cheaply and experiments confirm a few points. When mechanisms are unclear, designed experiments are more informative. We help weigh this choice, write down the decision criteria and plan checkpoints where the approach is reconsidered.

8. Uncertainty in process economics and sustainability

   Technoeconomic and sustainability assessments inherit uncertainty from yields, energy use, prices and equipment costs. Sensitivity analysis and Monte Carlo sampling show how conclusions change across plausible ranges. Reports state the basis year, scale and system boundaries, and avoid presenting a single cost or carbon figure as precise. Laboratory-scale performance generally underestimates problems that appear at industrial scale, so such assessments are presented as early-stage indications and not as investment-grade estimates.

Common pitfalls this service helps avoid

• Correlated kinetic parameters fitted separately. Individual estimates can be unreliable even if predictions fit well.
• Ignoring non-ideal flow. Real reactors depart from ideal models, affecting conversion and selectivity.
• Wrong thermodynamic model. Poor property predictions invalidate separation and equilibrium results.
• Extrapolating laboratory kinetics. Rate laws may change mechanism outside tested conditions.

What good reporting looks like

A process engineering report presents the system and chemistry, the data and measurement uncertainty, the balances and models, the parameter estimates with confidence regions, the validation, the optimisation setup and results, scale-up considerations and limitations. It does not constitute process safety assessment such as HAZOP or relief design, which requires qualified process safety engineers.

An illustrative scenario

Hypothetical example written for explanation. It is not a client case, a result or a testimonial.

Imagine a researcher fitting kinetics for a catalytic reaction. A single activation energy and pre-exponential factor fit the data well, but confidence regions are elongated, showing strong correlation. Reparameterising at a reference temperature gives tighter estimates. Testing two candidate mechanisms shows the data cannot clearly discriminate between them, and the report recommends additional experiments at a higher temperature.

Questions to settle before you begin

• Do my balances close within measurement uncertainty?
• Are my kinetic parameters identifiable from my data?
• Which thermodynamic model is valid for my mixture?
• What flow non-idealities could affect my reactor?
• What would change on scale-up?

Key terms

CSTR

A continuous stirred-tank reactor with uniform composition.

Residence-time distribution

The distribution of times fluid elements spend in a reactor.

Damköhler number

The ratio of reaction rate to transport rate.

Data reconciliation

Adjusting measurements to satisfy balances.

Identifiability

Whether parameters can be uniquely determined from data.

Working together on Materials, Chemical & Packaging Technology projects

Materials, chemical and packaging engagements begin with the material system or process, the properties of interest and the test methods used, together with any standards that provide context. We ask for composition and processing descriptions, test data with conditions, and the performance targets. Analysis separates what the data show from what is inferred from literature, and it states the limits of transfer between lab and production scale. Reports describe uncertainty in measurements and avoid implying regulatory compliance, which requires formal testing and approval.

Few projects need a single service in isolation. These are the services most often combined with this one.

• Response Surface Methodology (RSM)

  Sequential experimentation and modelling to relate responses to several factors and locate favourable operating conditions, using central composite or Box-Behnken style designs and fitted quadratic surfaces.

• Multiphase Flow Simulation

  Simulation of flows with more than one phase, such as gas-liquid, liquid-liquid and particle-laden systems, where model choice (Eulerian, Lagrangian, interface tracking) strongly shapes the result.

• Heat Transfer & Thermal Simulation

  Conduction, convection and radiation analysis of components and systems, including conjugate heat transfer, transient heating and cooling, and thermal management studies.

• Engineering Optimisation

  Formulation and solution of optimisation problems in engineering design and operation: defining objectives, variables and constraints carefully, then choosing an algorithm suited to the problem structure.