How Town Planning Works | TPW-0268 — The Materials Recovery Facility Capacity Test: How Mixed Recyclables, Contamination, Sorting Lines, Batteries, Fire, Residue and Commodity Markets Become One Land-Use System

Town planning becomes difficult when an urban service looks like a single land use on a map but behaves like an interconnected operating system. A materials recovery facility (MRF) needs land, access, utilities, environmental controls, emergency capacity and dependable interfaces beyond its fence or corridor. If one of those elements is missing, a technically impressive project can still fail the city it is meant to serve.

The question is current rather than theoretical. UN-Habitat reported on 31 March 2026 that the Korogocho Material Recovery Facility in Nairobi had begun operating as a decentralised dry-recyclables sorting and processing facility linked to community livelihoods. The World Bank’s 2026 What a Waste 3.0 work places municipal solid-waste infrastructure and circularity at the centre of urban-service planning, while current city programmes continue to finance collection, transfer, treatment and recovery systems. APA’s 2026 Trend Report for Planners likewise asks planners to prepare for infrastructure and environmental change rather than freeze today’s assumptions into long-lived land-use decisions.

The reader job is precise: How should a planning authority decide whether a proposed materials recovery facility (MRF) is genuinely capable of doing its stated job, where is the first binding constraint, and what measurable conditions keep the system compatible as demand, technology, climate and surrounding land uses change? The case should be able to explain tonnes per hour, material composition, capture and contamination, recovery yield and product purity, residue, peak queue time, safe inventory, verified downstream outlets and a credible outage mode, rather than relying on labels such as resilient, circular, smart or strategic.

This article owns the facility-level dry-recyclables sorting and recovery problem. It does not replace the Circular Town’s citywide material-loop owner, the general waste-system owner, the E-Waste Recovery and Repair Hub, the Advanced Plastics Recovery Hub, Textile Recovery, Battery Recycling, Circular Construction Materials, transport, finance, government or civilisation. It asks whether one MRF can physically turn a defined incoming recyclable stream into verified products without creating a larger fire, traffic, residue, worker-safety or neighbourhood problem.

1. Start with the actual incoming stream, not the blue-bin label

A MRF cannot be sized from the phrase “mixed recyclables.” Characterise paper grades, cardboard, metals, rigid plastics, films, glass, food residue, liquids, batteries and non-target objects by mass, volume and season. For planning purposes, this issue should be translated into a measurable operating envelope instead of left as descriptive prose. The application should show where the constraint appears physically, what normal margin remains, how that margin changes at the design peak, and what happens when one credible component or access route is unavailable. A decision can then distinguish a short operational upset from a structural capacity failure. The same evidence should be usable after opening, so the monitoring indicator and the review trigger need to be defined before approval rather than invented after complaints or service breakdown.

Planning test: What mass and composition cross the scale house on an ordinary day, a peak day and a bad-contamination day?

2. Separate participation, set-out and capture rates

High household participation does not prove high facility tonnage. Distinguish participation, set-out frequency, kilograms per household, capture by material and diversion through deposits, commercial systems or drop-off sites. The hard cases occur when several modest pressures coincide rather than when one variable reaches an extreme in isolation. A useful stress test combines peak demand, maintenance, difficult weather and one plausible failure, then traces the consequence through the facility and its external interfaces. This avoids the false reassurance of nameplate capacity. It also forces the proponent to identify a safe degraded mode: what can continue, what must slow, what must divert elsewhere and which threshold requires shutdown. Planning value comes from knowing the first binding constraint and the time available to respond.

Planning test: Which behavioural variable changes facility load most, and is it measured rather than assumed?

3. Define the service catchment before designing the plant

The catchment determines composition, haul pattern, population-growth exposure and contractual dependency. Regional plants can become brittle if neighbouring jurisdictions change collection policy or divert valuable fractions. Evidence should be traceable to site data, engineering assumptions, sampling or a named operating rule. Where the value is uncertain, the uncertainty should be shown as a range rather than buried in one precise-looking figure. The range can then be connected to a trigger for additional capacity, a changed operating condition or a new review. This keeps the plan adaptable without turning ‘future flexibility’ into permission for an undersized facility. The authority should also be able to verify the claim independently through records, inspection or a commissioning test rather than relying on an operator’s internal judgement.

Planning test: Which jurisdictions and customer classes are committed to the plant, for how long, and under what tonnage assumptions?

4. Use hourly arrival profiles, not annual tonnes

Collection fleets often arrive in waves after route completion. A plant with adequate annual sorting capacity can still fail at the gate because weighing, tipping or bay turnover cannot absorb the peak hour. A strong plan identifies the handoff to neighbouring systems instead of claiming to solve everything within one boundary. The facility should not take ownership from transport, finance, government, geography or another established technical owner, but it must prove that the physical interface is real and has capacity. A road, downstream plant, buyer, pipe, power supply or network connection can become the actual bottleneck even when the on-site equipment is adequate. The planning test is therefore end-to-end enough to expose dependency, while remaining narrow enough to preserve the canonical owner of the adjacent system.

Planning test: Can every realistic peak-hour arrival be received without blocking the public road, footway, cycleway or neighbouring access?

5. Make the weighbridge part of the process model

The weighbridge is the first mass-balance, contamination and circulation control point. Too few lanes create queues; weak load data makes it impossible to reconcile inbound tonnes with products and residue. The relevant comparison is not perfection versus failure. It is the proposed arrangement versus realistic alternatives: another site, different phasing, better source control, a smaller first stage, a stronger buffer, a different route or a different operating rule. That comparison separates impacts that are inherent in the urban service from impacts caused by one design choice. It also prevents planning from accepting unnecessary land consumption or neighbourhood burden simply because the applicant has presented only one engineering scheme. The preferred option should be able to explain what it improves, what it worsens and which residual risk still needs management.

Planning test: Does the gate record enough information to explain where each tonne came from and where it ultimately went?

6. Treat the tipping floor as inventory, not empty space

Covered receiving space should be expressed in tonnes and hours, adjusted for bulk density, pile height and fire separation. A large floor is not unlimited buffer capacity. Monitoring should be designed around decisions, not around the desire to collect data. The authority needs a small number of indicators that correspond to the claimed benefit and the principal failure mode, with thresholds that prompt a specified response. A dashboard that cannot change operations is reporting overhead. Good indicators also help separate temporary noise in the data from a sustained decline in performance. Where community exposure is part of the issue, monitoring should make the spatial and temporal pattern visible enough to test whether a permit condition is actually protecting the intended receptor.

Planning test: How many hours of incoming material can the tipping floor safely hold if the sorting line stops at the busiest time?

7. Decide whether bags are a collection convenience or a processing penalty

Bagged recyclables may help households but create opening, film and tangling problems. Loose collection creates different litter risks. Plant design must follow the actual collection system. Spatial design matters because operating assumptions eventually become land-use effects. A queue becomes a road hazard, excess inventory becomes a fire load, an uncontrolled opening becomes an odour path, a low electrical room becomes a flood vulnerability, and a shared corridor becomes a common point of failure. The plan should therefore connect every important process assumption to a place on the site or route. This makes later changes easier to evaluate: planners can see whether a new bay, tenant, pipe, duct or storage area stays inside the approved envelope or silently moves the risk toward another receptor.

Planning test: What equipment and labour step deals with every bag that reaches the line, and where does removed film go?

8. Design pre-sort as a safety barrier

Large, dangerous or tangling objects should be intercepted before high-speed equipment. Cylinders, lithium batteries, hoses, scrap, textiles and bulky items need visible reject routes and safe worker access. Future change should be treated as a bounded scenario rather than a blank cheque. The design can reserve land, utility headroom, alternative connections, replaceable equipment or modular expansion while still stating today’s permitted load. Review thresholds are especially valuable where technology, material streams, climate conditions, regulation or surrounding development may change over a long asset life. A flexible plan is not one with no limits; it is one that knows which parts can change safely, which assumptions are structural and what evidence is required before the next stage is activated.

Planning test: Which objects must be intercepted before the first major machine, and can workers do that without entering an uncontrolled hazard zone?

9. Match screens to the real size distribution

Disc, star, trommel and ballistic screens separate partly by size and shape. Broken glass, flattened containers and compacted feed can behave very differently from vendor demonstration material. For planning purposes, this issue should be translated into a measurable operating envelope instead of left as descriptive prose. The application should show where the constraint appears physically, what normal margin remains, how that margin changes at the design peak, and what happens when one credible component or access route is unavailable. A decision can then distinguish a short operational upset from a structural capacity failure. The same evidence should be usable after opening, so the monitoring indicator and the review trigger need to be defined before approval rather than invented after complaints or service breakdown.

Planning test: What fraction is each screen intended to create, and what contamination appears when feed size distribution shifts?

10. Treat optical sorters as classifiers with limits

Optical systems depend on belt loading, singulation, surface condition, colour and target polymer or fibre. Misses, false positives and recirculation must be included in throughput claims. The hard cases occur when several modest pressures coincide rather than when one variable reaches an extreme in isolation. A useful stress test combines peak demand, maintenance, difficult weather and one plausible failure, then traces the consequence through the facility and its external interfaces. This avoids the false reassurance of nameplate capacity. It also forces the proponent to identify a safe degraded mode: what can continue, what must slow, what must divert elsewhere and which threshold requires shutdown. Planning value comes from knowing the first binding constraint and the time available to respond.

Planning test: At design belt loading, what recovery and purity can the system sustain for hours rather than minutes?

11. Use magnets and eddy-current systems where physics helps

Ferrous and non-ferrous metal recovery can protect equipment and produce valuable outputs, but placement, nested objects, foil and composites affect results. Evidence should be traceable to site data, engineering assumptions, sampling or a named operating rule. Where the value is uncertain, the uncertainty should be shown as a range rather than buried in one precise-looking figure. The range can then be connected to a trigger for additional capacity, a changed operating condition or a new review. This keeps the plan adaptable without turning ‘future flexibility’ into permission for an undersized facility. The authority should also be able to verify the claim independently through records, inspection or a commissioning test rather than relying on an operator’s internal judgement.

Planning test: Which metal fraction is targeted at each stage, and how is product purity checked before sale?

12. Give glass its own damage-and-quality strategy

Broken glass is dense, abrasive and capable of contaminating paper and other products. Count it as recovery only when a defined quality and downstream use exist. A strong plan identifies the handoff to neighbouring systems instead of claiming to solve everything within one boundary. The facility should not take ownership from transport, finance, government, geography or another established technical owner, but it must prove that the physical interface is real and has capacity. A road, downstream plant, buyer, pipe, power supply or network connection can become the actual bottleneck even when the on-site equipment is adequate. The planning test is therefore end-to-end enough to expose dependency, while remaining narrow enough to preserve the canonical owner of the adjacent system.

Planning test: What specification and end market make recovered glass a product rather than a heavy residue stream?

13. Protect paper from moisture and fines

Wet, food-soiled fibre or paper contaminated by glass fines loses value quickly. Covered receiving, drainage and process sequence are therefore quality infrastructure, not cosmetic features. The relevant comparison is not perfection versus failure. It is the proposed arrangement versus realistic alternatives: another site, different phasing, better source control, a smaller first stage, a stronger buffer, a different route or a different operating rule. That comparison separates impacts that are inherent in the urban service from impacts caused by one design choice. It also prevents planning from accepting unnecessary land consumption or neighbourhood burden simply because the applicant has presented only one engineering scheme. The preferred option should be able to explain what it improves, what it worsens and which residual risk still needs management.

Planning test: What moisture and prohibitive-material limit applies to each fibre grade, and how often is it expected to be exceeded?

14. Do not treat all plastics as one commodity

PET, HDPE, polypropylene, films, coloured resin and multilayer packaging have different sorting behaviour and markets. State which grades the MRF can reliably produce. Monitoring should be designed around decisions, not around the desire to collect data. The authority needs a small number of indicators that correspond to the claimed benefit and the principal failure mode, with thresholds that prompt a specified response. A dashboard that cannot change operations is reporting overhead. Good indicators also help separate temporary noise in the data from a sustained decline in performance. Where community exposure is part of the issue, monitoring should make the spatial and temporal pattern visible enough to test whether a permit condition is actually protecting the intended receptor.

Planning test: Which plastic grade leaves under a saleable specification, and which merely leaves under another waste code?

15. Design for films and tanglers before they stop the line

Film, straps, hoses, cords and textiles wrap around shafts and screens, causing dangerous cleaning and downtime. The plant should include realistic lost-production time and maintenance access. Spatial design matters because operating assumptions eventually become land-use effects. A queue becomes a road hazard, excess inventory becomes a fire load, an uncontrolled opening becomes an odour path, a low electrical room becomes a flood vulnerability, and a shared corridor becomes a common point of failure. The plan should therefore connect every important process assumption to a place on the site or route. This makes later changes easier to evaluate: planners can see whether a new bay, tenant, pipe, duct or storage area stays inside the approved envelope or silently moves the risk toward another receptor.

Planning test: How many minutes of line availability are lost to wrapping and cleaning in the design case?

16. Treat lithium batteries as a fire pathway through the whole facility

Small batteries can enter recycling inside toys, electronics or loose cells. Crushing and impact can trigger thermal runaway, so detection, quarantine, suppression and fire access must cover every stage. Future change should be treated as a bounded scenario rather than a blank cheque. The design can reserve land, utility headroom, alternative connections, replaceable equipment or modular expansion while still stating today’s permitted load. Review thresholds are especially valuable where technology, material streams, climate conditions, regulation or surrounding development may change over a long asset life. A flexible plan is not one with no limits; it is one that knows which parts can change safely, which assumptions are structural and what evidence is required before the next stage is activated.

Planning test: Where can an undetected battery travel, and what prevents each stage from becoming the ignition point for a large fuel load?

17. Plan for pressurised containers and hidden hazardous items

Aerosols, gas cylinders, chemicals and sharps arrive despite collection rules. Isolation and lawful downstream management should be designed rather than improvised after an incident. For planning purposes, this issue should be translated into a measurable operating envelope instead of left as descriptive prose. The application should show where the constraint appears physically, what normal margin remains, how that margin changes at the design peak, and what happens when one credible component or access route is unavailable. A decision can then distinguish a short operational upset from a structural capacity failure. The same evidence should be usable after opening, so the monitoring indicator and the review trigger need to be defined before approval rather than invented after complaints or service breakdown.

Planning test: What happens in the first ten minutes after staff find a cylinder, chemical container or suspicious object?

18. Measure organics contamination as a process burden

Food and liquids increase odour, pests, moisture and worker exposure while degrading fibre and container quality. Weight percentage alone can understate operational harm. The hard cases occur when several modest pressures coincide rather than when one variable reaches an extreme in isolation. A useful stress test combines peak demand, maintenance, difficult weather and one plausible failure, then traces the consequence through the facility and its external interfaces. This avoids the false reassurance of nameplate capacity. It also forces the proponent to identify a safe degraded mode: what can continue, what must slow, what must divert elsewhere and which threshold requires shutdown. Planning value comes from knowing the first binding constraint and the time available to respond.

Planning test: Which contaminants create the greatest downtime or quality penalty per tonne, and how does collection policy respond?

19. Design manual sorting around human reach and attention

Belt speed, working height, reach, lighting, heat, noise and task rotation determine whether workers can actually maintain required quality over a shift. Evidence should be traceable to site data, engineering assumptions, sampling or a named operating rule. Where the value is uncertain, the uncertainty should be shown as a range rather than buried in one precise-looking figure. The range can then be connected to a trigger for additional capacity, a changed operating condition or a new review. This keeps the plan adaptable without turning ‘future flexibility’ into permission for an undersized facility. The authority should also be able to verify the claim independently through records, inspection or a commissioning test rather than relying on an operator’s internal judgement.

Planning test: Can a worker sustain the required pick rate without unsafe reaching, excessive noise exposure or fatigue-driven mistakes?

20. Control dust as both a health and fire issue

Dry fibre, fines and broken material generate airborne dust. Extraction, enclosure, housekeeping and ignition control should be designed together. A strong plan identifies the handoff to neighbouring systems instead of claiming to solve everything within one boundary. The facility should not take ownership from transport, finance, government, geography or another established technical owner, but it must prove that the physical interface is real and has capacity. A road, downstream plant, buyer, pipe, power supply or network connection can become the actual bottleneck even when the on-site equipment is adequate. The planning test is therefore end-to-end enough to expose dependency, while remaining narrow enough to preserve the canonical owner of the adjacent system.

Planning test: Where is dust generated, captured, cleaned and disposed, and what threshold triggers additional controls?

21. Map noise by source and time of day

Reversing alarms, compactors, balers, glass handling, loaders, fans and trucks create different sound profiles; dawn impulses can matter more than a steady daytime average. The relevant comparison is not perfection versus failure. It is the proposed arrangement versus realistic alternatives: another site, different phasing, better source control, a smaller first stage, a stronger buffer, a different route or a different operating rule. That comparison separates impacts that are inherent in the urban service from impacts caused by one design choice. It also prevents planning from accepting unnecessary land consumption or neighbourhood burden simply because the applicant has presented only one engineering scheme. The preferred option should be able to explain what it improves, what it worsens and which residual risk still needs management.

Planning test: Which source controls the night or early-morning receptor limit, and can the site operate if that source is restricted?

22. Design the fire compartment before choosing suppression

Mixed piles and baled plastics or paper can sustain difficult fires. Pile limits, separation, detection, access, water supply and moveable inventory matter as much as suppression technology. Monitoring should be designed around decisions, not around the desire to collect data. The authority needs a small number of indicators that correspond to the claimed benefit and the principal failure mode, with thresholds that prompt a specified response. A dashboard that cannot change operations is reporting overhead. Good indicators also help separate temporary noise in the data from a sustained decline in performance. Where community exposure is part of the issue, monitoring should make the spatial and temporal pattern visible enough to test whether a permit condition is actually protecting the intended receptor.

Planning test: What is the maximum combustible inventory in each fire area, and who can stop intake before it is exceeded?

23. Keep contaminated firewater out of drains and waterways

A major fire can generate polluted water containing ash, plastics, metals and chemicals. Grading, isolation valves and containment must handle the design fire plus adverse weather. Spatial design matters because operating assumptions eventually become land-use effects. A queue becomes a road hazard, excess inventory becomes a fire load, an uncontrolled opening becomes an odour path, a low electrical room becomes a flood vulnerability, and a shared corridor becomes a common point of failure. The plan should therefore connect every important process assumption to a place on the site or route. This makes later changes easier to evaluate: planners can see whether a new bay, tenant, pipe, duct or storage area stays inside the approved envelope or silently moves the risk toward another receptor.

Planning test: Where does firewater go when the largest credible fire occurs during heavy rain?

24. Separate process drainage from clean stormwater

Tipping floors, washdown areas and waste-contact yards should not drain like ordinary roofs. The clean/dirty drainage boundary must be visible and maintainable. Future change should be treated as a bounded scenario rather than a blank cheque. The design can reserve land, utility headroom, alternative connections, replaceable equipment or modular expansion while still stating today’s permitted load. Review thresholds are especially valuable where technology, material streams, climate conditions, regulation or surrounding development may change over a long asset life. A flexible plan is not one with no limits; it is one that knows which parts can change safely, which assumptions are structural and what evidence is required before the next stage is activated.

Planning test: Can an inspector trace every floor drain and yard inlet to its receiving system without relying on institutional memory?

25. Treat compressed air, power and controls as production utilities

Sorting depends on electricity, compressed air, sensors, conveyors and control systems. A short voltage event can stop the line and rapidly create inbound backlog. For planning purposes, this issue should be translated into a measurable operating envelope instead of left as descriptive prose. The application should show where the constraint appears physically, what normal margin remains, how that margin changes at the design peak, and what happens when one credible component or access route is unavailable. A decision can then distinguish a short operational upset from a structural capacity failure. The same evidence should be usable after opening, so the monitoring indicator and the review trigger need to be defined before approval rather than invented after complaints or service breakdown.

Planning test: Which utility failure stops acceptance first, and how long can the site operate before public collection is disrupted?

26. Design an outage mode before opening day

Every MRF experiences maintenance and unplanned stoppage. Diversion, safe buffering, extended shifts or reciprocal processing should be arranged before the first breakdown. The hard cases occur when several modest pressures coincide rather than when one variable reaches an extreme in isolation. A useful stress test combines peak demand, maintenance, difficult weather and one plausible failure, then traces the consequence through the facility and its external interfaces. This avoids the false reassurance of nameplate capacity. It also forces the proponent to identify a safe degraded mode: what can continue, what must slow, what must divert elsewhere and which threshold requires shutdown. Planning value comes from knowing the first binding constraint and the time available to respond.

Planning test: What happens to tomorrow morning’s collection routes if the main line becomes unavailable tonight?

27. Size balers around outbound logistics, not only line speed

Bale density, dimensions, strapping and production must match forklifts, storage, trucks, containers and buyers. Baling bottlenecks can turn sorted product into a fire inventory. Evidence should be traceable to site data, engineering assumptions, sampling or a named operating rule. Where the value is uncertain, the uncertainty should be shown as a range rather than buried in one precise-looking figure. The range can then be connected to a trigger for additional capacity, a changed operating condition or a new review. This keeps the plan adaptable without turning ‘future flexibility’ into permission for an undersized facility. The authority should also be able to verify the claim independently through records, inspection or a commissioning test rather than relying on an operator’s internal judgement.

Planning test: Can the balers process peak recovered output while one unit is unavailable for maintenance?

28. Limit bale storage by market and fire reality

Inventory rises when buyers slow, prices fall or transport is disrupted. Storage should be limited by days, material class and fire separation rather than whatever yard remains empty. A strong plan identifies the handoff to neighbouring systems instead of claiming to solve everything within one boundary. The facility should not take ownership from transport, finance, government, geography or another established technical owner, but it must prove that the physical interface is real and has capacity. A road, downstream plant, buyer, pipe, power supply or network connection can become the actual bottleneck even when the on-site equipment is adequate. The planning test is therefore end-to-end enough to expose dependency, while remaining narrow enough to preserve the canonical owner of the adjacent system.

Planning test: How many days of product can the site store before intake or production must be reduced?

29. Prove every outbound route with a specification and receiver

Each recovered stream should have a product specification, named receiver or processor, transport route and rejection procedure. A single recycling percentage hides too much. The relevant comparison is not perfection versus failure. It is the proposed arrangement versus realistic alternatives: another site, different phasing, better source control, a smaller first stage, a stronger buffer, a different route or a different operating rule. That comparison separates impacts that are inherent in the urban service from impacts caused by one design choice. It also prevents planning from accepting unnecessary land consumption or neighbourhood burden simply because the applicant has presented only one engineering scheme. The preferred option should be able to explain what it improves, what it worsens and which residual risk still needs management.

Planning test: For each output, who receives it, what specification applies, and what is the lawful fallback if the load is refused?

30. Model commodity downturns as operating scenarios

Paper, metal and plastic prices fluctuate. Low prices can lengthen storage, weaken quality control and increase residue even when technical equipment is unchanged. Monitoring should be designed around decisions, not around the desire to collect data. The authority needs a small number of indicators that correspond to the claimed benefit and the principal failure mode, with thresholds that prompt a specified response. A dashboard that cannot change operations is reporting overhead. Good indicators also help separate temporary noise in the data from a sustained decline in performance. Where community exposure is part of the issue, monitoring should make the spatial and temporal pattern visible enough to test whether a permit condition is actually protecting the intended receptor.

Planning test: Can the plant remain within storage, fire and residue limits during a sustained low-price period?

31. Give residue a real destination

Sorting inevitably creates fines, non-targets and unrecoverable material. The mass balance should identify typical and worst-case residue shares and lawful destination capacity. Spatial design matters because operating assumptions eventually become land-use effects. A queue becomes a road hazard, excess inventory becomes a fire load, an uncontrolled opening becomes an odour path, a low electrical room becomes a flood vulnerability, and a shared corridor becomes a common point of failure. The plan should therefore connect every important process assumption to a place on the site or route. This makes later changes easier to evaluate: planners can see whether a new bay, tenant, pipe, duct or storage area stays inside the approved envelope or silently moves the risk toward another receptor.

Planning test: If residue rises by ten percentage points because incoming quality worsens, can the facility still manage it lawfully?

32. Reconcile the mass balance at the scale house

Inbound mass should approximately reconcile with products, transfers, residue, inventory movement and moisture change. Unexplained gaps undermine recovery claims. Future change should be treated as a bounded scenario rather than a blank cheque. The design can reserve land, utility headroom, alternative connections, replaceable equipment or modular expansion while still stating today’s permitted load. Review thresholds are especially valuable where technology, material streams, climate conditions, regulation or surrounding development may change over a long asset life. A flexible plan is not one with no limits; it is one that knows which parts can change safely, which assumptions are structural and what evidence is required before the next stage is activated.

Planning test: Can the operator explain a full month of inbound and outbound mass within a defined reconciliation tolerance?

33. Distinguish recovery rate from quality-adjusted recovery

A high captured tonnage is weak evidence if bales are routinely rejected or re-sorted elsewhere. Track whether outputs actually meet their intended specification. For planning purposes, this issue should be translated into a measurable operating envelope instead of left as descriptive prose. The application should show where the constraint appears physically, what normal margin remains, how that margin changes at the design peak, and what happens when one credible component or access route is unavailable. A decision can then distinguish a short operational upset from a structural capacity failure. The same evidence should be usable after opening, so the monitoring indicator and the review trigger need to be defined before approval rather than invented after complaints or service breakdown.

Planning test: How much material reaches a verified next use without being rejected, downgraded or disposed after leaving the gate?

34. Integrate informal and community recovery without romanticising unsafe work

Where informal recovery already exists, formalisation can improve safety and market access or displace livelihoods. Existing actors should be mapped as part of the real material system. The hard cases occur when several modest pressures coincide rather than when one variable reaches an extreme in isolation. A useful stress test combines peak demand, maintenance, difficult weather and one plausible failure, then traces the consequence through the facility and its external interfaces. This avoids the false reassurance of nameplate capacity. It also forces the proponent to identify a safe degraded mode: what can continue, what must slow, what must divert elsewhere and which threshold requires shutdown. Planning value comes from knowing the first binding constraint and the time available to respond.

Planning test: Who currently earns income from this material stream, and what changes for those people when the MRF opens?

35. Test cumulative burden at the receiving neighbourhood

A compliant MRF can still join depots, highways and waste uses already concentrated in one district. Truck, noise, dust and fire burdens should be assessed cumulatively. Evidence should be traceable to site data, engineering assumptions, sampling or a named operating rule. Where the value is uncertain, the uncertainty should be shown as a range rather than buried in one precise-looking figure. The range can then be connected to a trigger for additional capacity, a changed operating condition or a new review. This keeps the plan adaptable without turning ‘future flexibility’ into permission for an undersized facility. The authority should also be able to verify the claim independently through records, inspection or a commissioning test rather than relying on an operator’s internal judgement.

Planning test: What existing burdens surround the site, and does the MRF add materially to the same exposure pathways?

36. Use buffers to solve specific effects, not to hide weak design

Distance can reduce impact, but an oversized setback is not a substitute for enclosure, filtration, traffic control or fire design. Buffers should correspond to a named receptor and effect. A strong plan identifies the handoff to neighbouring systems instead of claiming to solve everything within one boundary. The facility should not take ownership from transport, finance, government, geography or another established technical owner, but it must prove that the physical interface is real and has capacity. A road, downstream plant, buyer, pipe, power supply or network connection can become the actual bottleneck even when the on-site equipment is adequate. The planning test is therefore end-to-end enough to expose dependency, while remaining narrow enough to preserve the canonical owner of the adjacent system.

Planning test: Which measured effect requires each buffer, and could better engineering reduce the land consumed?

37. Orient buildings and doors around receptors

Tipping doors, exhausts, truck aprons and bale yards determine where odour, dust and noise travel. Building mass and orientation are therefore environmental controls. The relevant comparison is not perfection versus failure. It is the proposed arrangement versus realistic alternatives: another site, different phasing, better source control, a smaller first stage, a stronger buffer, a different route or a different operating rule. That comparison separates impacts that are inherent in the urban service from impacts caused by one design choice. It also prevents planning from accepting unnecessary land consumption or neighbourhood burden simply because the applicant has presented only one engineering scheme. The preferred option should be able to explain what it improves, what it worsens and which residual risk still needs management.

Planning test: With prevailing winds and sensitive receptors shown, are the dirtiest openings facing the least harmful direction?

38. Reserve space for a second line without pre-building it

Future capacity can be protected through land, utility, conveyor and circulation reservations while phase one remains independently operable. Monitoring should be designed around decisions, not around the desire to collect data. The authority needs a small number of indicators that correspond to the claimed benefit and the principal failure mode, with thresholds that prompt a specified response. A dashboard that cannot change operations is reporting overhead. Good indicators also help separate temporary noise in the data from a sustained decline in performance. Where community exposure is part of the issue, monitoring should make the spatial and temporal pattern visible enough to test whether a permit condition is actually protecting the intended receptor.

Planning test: Can throughput be materially increased without shutting the first line or rebuilding the gate, utilities and fire system?

39. Plan for changes in packaging composition

Light-weighting, deposits, fibre substitution and flexible packaging change both tonnes and bulk volume. A MRF optimised for yesterday’s mix can become obsolete while total recycling remains high. Spatial design matters because operating assumptions eventually become land-use effects. A queue becomes a road hazard, excess inventory becomes a fire load, an uncontrolled opening becomes an odour path, a low electrical room becomes a flood vulnerability, and a shared corridor becomes a common point of failure. The plan should therefore connect every important process assumption to a place on the site or route. This makes later changes easier to evaluate: planners can see whether a new bay, tenant, pipe, duct or storage area stays inside the approved envelope or silently moves the risk toward another receptor.

Planning test: Which material trend would make today’s equipment configuration obsolete fastest, and what can be changed modularly?

40. Treat extreme heat as an operating constraint

Heat affects workers, equipment, odour and fire risk. Heat-specific operating limits should be explicit if reduced line speed or additional ventilation lowers throughput. Future change should be treated as a bounded scenario rather than a blank cheque. The design can reserve land, utility headroom, alternative connections, replaceable equipment or modular expansion while still stating today’s permitted load. Review thresholds are especially valuable where technology, material streams, climate conditions, regulation or surrounding development may change over a long asset life. A flexible plan is not one with no limits; it is one that knows which parts can change safely, which assumptions are structural and what evidence is required before the next stage is activated.

Planning test: What changes on the hottest design day, and can collection service continue without unsafe worker exposure?

41. Keep the site functional during intense rainfall and flood

Low-lying waste sites can lose access, spread litter or contaminate floodwater. Finished levels, drainage and electrical placement should use future hazard, not only historic rainfall. For planning purposes, this issue should be translated into a measurable operating envelope instead of left as descriptive prose. The application should show where the constraint appears physically, what normal margin remains, how that margin changes at the design peak, and what happens when one credible component or access route is unavailable. A decision can then distinguish a short operational upset from a structural capacity failure. The same evidence should be usable after opening, so the monitoring indicator and the review trigger need to be defined before approval rather than invented after complaints or service breakdown.

Planning test: What flood depth or access closure triggers shutdown, and where does incoming material go during that period?

42. Design emergency access for the largest credible incident

Fire appliances need access even when trucks, bales or damaged equipment occupy the yard. The emergency diagram should assume real obstruction, not an empty site. The hard cases occur when several modest pressures coincide rather than when one variable reaches an extreme in isolation. A useful stress test combines peak demand, maintenance, difficult weather and one plausible failure, then traces the consequence through the facility and its external interfaces. This avoids the false reassurance of nameplate capacity. It also forces the proponent to identify a safe degraded mode: what can continue, what must slow, what must divert elsewhere and which threshold requires shutdown. Planning value comes from knowing the first binding constraint and the time available to respond.

Planning test: Can responders reach both sides of the largest combustible inventory while normal circulation is partly blocked?

43. Make permits measurable enough to enforce

Useful conditions define hours, inventory, enclosure, traffic, firewater, accepted feed, monitoring and reporting rather than vague duties to minimise nuisance. Evidence should be traceable to site data, engineering assumptions, sampling or a named operating rule. Where the value is uncertain, the uncertainty should be shown as a range rather than buried in one precise-looking figure. The range can then be connected to a trigger for additional capacity, a changed operating condition or a new review. This keeps the plan adaptable without turning ‘future flexibility’ into permission for an undersized facility. The authority should also be able to verify the claim independently through records, inspection or a commissioning test rather than relying on an operator’s internal judgement.

Planning test: Which five operating variables most strongly protect neighbours and safety, and can they be verified from records or inspection?

44. Use commissioning to prove the promised throughput

Nameplate machine capacity does not prove whole-plant performance. Commissioning should use representative feed over sustained hours and measure recovery, purity, residue, queue, dust and noise. A strong plan identifies the handoff to neighbouring systems instead of claiming to solve everything within one boundary. The facility should not take ownership from transport, finance, government, geography or another established technical owner, but it must prove that the physical interface is real and has capacity. A road, downstream plant, buyer, pipe, power supply or network connection can become the actual bottleneck even when the on-site equipment is adequate. The planning test is therefore end-to-end enough to expose dependency, while remaining narrow enough to preserve the canonical owner of the adjacent system.

Planning test: What sustained performance test must pass before the facility accepts its full service-area load?

45. Monitor inbound quality by route or customer

Average contamination can conceal one district or customer causing repeated problems. Granular feedback allows targeted education, enforcement or collection redesign. The relevant comparison is not perfection versus failure. It is the proposed arrangement versus realistic alternatives: another site, different phasing, better source control, a smaller first stage, a stronger buffer, a different route or a different operating rule. That comparison separates impacts that are inherent in the urban service from impacts caused by one design choice. It also prevents planning from accepting unnecessary land consumption or neighbourhood burden simply because the applicant has presented only one engineering scheme. The preferred option should be able to explain what it improves, what it worsens and which residual risk still needs management.

Planning test: Can the operator identify source patterns behind contamination and feed that information back upstream?

46. Track downtime as a land-use performance indicator

When the line stops, trucks queue, tipping piles grow and diversion haulage rises. Downtime therefore has spatial consequences beyond factory productivity. Monitoring should be designed around decisions, not around the desire to collect data. The authority needs a small number of indicators that correspond to the claimed benefit and the principal failure mode, with thresholds that prompt a specified response. A dashboard that cannot change operations is reporting overhead. Good indicators also help separate temporary noise in the data from a sustained decline in performance. Where community exposure is part of the issue, monitoring should make the spatial and temporal pattern visible enough to test whether a permit condition is actually protecting the intended receptor.

Planning test: How many hours of unplanned downtime can occur before safe storage is exceeded or diversion becomes necessary?

47. Audit downstream destinations rather than trusting labels

A broker or exporter is not automatically final recovery. Periodically verify destination, treatment route and rejection rate for major output streams. Spatial design matters because operating assumptions eventually become land-use effects. A queue becomes a road hazard, excess inventory becomes a fire load, an uncontrolled opening becomes an odour path, a low electrical room becomes a flood vulnerability, and a shared corridor becomes a common point of failure. The plan should therefore connect every important process assumption to a place on the site or route. This makes later changes easier to evaluate: planners can see whether a new bay, tenant, pipe, duct or storage area stays inside the approved envelope or silently moves the risk toward another receptor.

Planning test: Does the authority know the next physical processing step for each major output rather than only the name of an intermediary?

48. Keep collection policy and plant design aligned

Accepted materials, cart type, frequency and compaction determine the feed. Changes should pass through a defined technical change-control process before they reshape facility performance. Future change should be treated as a bounded scenario rather than a blank cheque. The design can reserve land, utility headroom, alternative connections, replaceable equipment or modular expansion while still stating today’s permitted load. Review thresholds are especially valuable where technology, material streams, climate conditions, regulation or surrounding development may change over a long asset life. A flexible plan is not one with no limits; it is one that knows which parts can change safely, which assumptions are structural and what evidence is required before the next stage is activated.

Planning test: What testing and approval occurs before the collection system introduces a new material or packaging class?

49. Plan closure and successor use from the beginning

A MRF building may outlive its sorting line. Closure should address inventory, damaged or contaminated areas, equipment removal and a feasible successor industrial use. For planning purposes, this issue should be translated into a measurable operating envelope instead of left as descriptive prose. The application should show where the constraint appears physically, what normal margin remains, how that margin changes at the design peak, and what happens when one credible component or access route is unavailable. A decision can then distinguish a short operational upset from a structural capacity failure. The same evidence should be usable after opening, so the monitoring indicator and the review trigger need to be defined before approval rather than invented after complaints or service breakdown.

Planning test: If recovery technology moves elsewhere in twenty years, can the site be safely cleared and reused without hidden public liabilities?

50. Issue a capacity certificate, not a sustainability slogan

The final record should state feed assumptions, permitted throughput, peak receiving rate, inventory, recovery and residue routes, fire controls, outage strategy and expansion triggers. The hard cases occur when several modest pressures coincide rather than when one variable reaches an extreme in isolation. A useful stress test combines peak demand, maintenance, difficult weather and one plausible failure, then traces the consequence through the facility and its external interfaces. This avoids the false reassurance of nameplate capacity. It also forces the proponent to identify a safe degraded mode: what can continue, what must slow, what must divert elsewhere and which threshold requires shutdown. Planning value comes from knowing the first binding constraint and the time available to respond.

Planning test: Can the approving authority describe, in one page of measurable conditions, the envelope within which the MRF is genuinely fit for service?