Over the past few years, UK materials recycling facilities (MRFs) have come under mounting economic pressure caused by increased recyclate quality requirements and low recyclate price, which has forced operators to scrutinise operational expenditure.
Brexit looming may also cause additional pressure where the restricted freedom of movement could cut off the flow of more competitively priced labour that our, and other industries, have benefited from for many years. For example, Costa Coffee recently published an article in The Guardian to warn that wages will be pushed up if Brexit shrinks the pool of workers available to it.
Most modern MRFs currently work in a two- or three-shift system operating 18-22 hours per day, Monday to Saturday. As a result, staffing requirements are high with labour costs typically making up about 50% of the OPEX for a semi-automated MRF in the form of pickers, who are used to recover recyclable material manually and perform quality control tasks.
If we look at a typical modern UK MRF built in the past four to five years for processing comingled recyclates with glass, it will normally process around 20-25 tonnes per hour (tph) on one or two process lines. This line will have disc type screens or trommels, a ballistic separator and three to four NIR sorters to recover plastics and clean up the paper grades in a single pass, with some reliance on manual sorting to carry out the final quality control on recovered recyclates.
Regularly working with operators who are looking to improve MRF performance, a spokesperson from Dutch MRF manufacture Bollegraaf says: “The typical questions we are getting from existing plant operators are: ‘Do we automate? Do we try and push more through the plant? The material is getting more contaminated, so how do we clean it up more cheaply?’” So, with rising costs being a key factor in the operation of a medium-scaled MRF, does scaling up have advantages?
A typical modern MRF, as illustrated above, will cost around £6m. If the MRF is operational for 20 hours per day, 50 weeks of the year with 90% availability, the plant will operate for approximately 5,400 hours per year. Assuming a simple straight line depreciation of 12 years on the initial capital expenditure, the CAPEX depreciation per tonne will be £4.44.
Doubling the size of the MRF to increase capacity to 50tph would not reduce your CAPEX per tonne, as twice the amount of the same equipment would be required.
However, what if you increased the throughput of a single-line MRF processing comingled material with glass to 35tph? The main limitation of many MRFs is around the screening equipment at the start of the process. The throughput of trommel screens used to separate out the glass fraction and prepare the fibre and container fractions for NIR sorting, is limited by the fact that the actual screening of material only takes place where the material is in contact with the screening drum. As a result, even the largest trommel screen will only effectively screen up to 22-25tph, after which two lines are required.
One solution to this is to use driven disc type screens which ensure higher throughput. The advent of large 4m-wide screens, equipped with large diameter anti-wrapping shafts and larger diameter stars, has meant that the need for cleaning has been reduced throughout.
Observations of a MRF, based in Dallas, Texas, equipped with these new type screens revealed that they only required cleaning once per day, despite contamination levels of 16.4% and, of this, a high percentage was film. This particular plant processes a typical comingled infeed with glass at 35tph on a single process line. The plant cost around US$8m or £6.5m, bringing the CAPEX down to just £2.86 per tonne.
The key point is that when increasing the overall throughput of a MRF, just duplicating the number of lines will not affect the CAPEX per tonne unless the throughout per line is also increased, and this can only be done by using the correct equipment to ensure a high throughput and performance.
Based on the typical 25tph UK MRF model above with some NIR automation, it may
have around 17-19 sorters per shift used for pre-sorting (3), QC of paper grades (6-8) and plastic/metals (4-5) and recovery (2).
This would require 51-57 workers per day (three shifts) who are paid approximately £9.75 per hour each (including NI/holiday/pension costs) which would equate to costs per tonne of around £6.73-£7.41 for manual sorters.
According to Bollegraaf, you will always need the human touch when it comes to the final quality control of materials in an MRF. However, if you consider the scenario where a MRF had four additional NIR sorting to perform the main picking function with only one or two staff per product to perform the final QC/polishing of the material, there will be relative costs to consider:
- CAPEX would increase from £6m to approximately £7.5m (with the operating criteria being the same as above)
- The CAPEX depreciation per tonne would increase from £4.44 per tonne to £5.56 per tonne
- The number of sorters would be reduced to 13 for pre-sorting (3), QC of paper grades (4), plastic/metals (3) and recovery (2)
- Manual sorters would be reduced from 51-57 to approximately 39 per day with costs for manual sorters being around £5.07 per tonne
- The increase of CAPEX is £1.12 per tonne against an OPEX saving of up to £2.34 per tonne.
This is a simple illustration and there are additional key factors which will influence the decision to further automate a MRF or not, and these include:
- Availability of labour
- Recyclate quality requirements
- Space in existing MRF to retro fit new equipment
- Availability of CAPEX.
Let’s look again at the savings to be made if you go large and consider mega MRFs to process 60-70tph of comingled materials with glass.
As illustrated above, purely increasing the number of lines of a typical 25tph MRF is not sufficient and will not bring any overall savings. To achieve real economies of scale
you will need to ensure that each process line increases its throughput compared to its current position, i.e. from 20-25tph to 30-35tph per line
Bollegraaf has been very fortunate to have built 16 of the 19 mega MRFs (processing greater than 50tph) in the USA. These MRFs have two process lines with a capacity of 30-35tph each and, based on their operating data, the following data can be used to illustrate the mega MRF concept.
The cost of a mega MRF with some automation ranges from US$13-16m or £11-13m with NIRs for the plastics recovery. Based on the operating criteria above, where the MRF is operational for 20 hours per day, 50 weeks of the year with 90% availability, the plant will operate for approximately 5,400 hours per year. If we assume a simple straight line depreciation of 12 years on the initial capital expenditure, the CAPEX depreciation will be £3.08-£3.85 per tonne. At 70tph this would reduce further to £2.64-£3.31 per tonne.
However, US MRFs generally rely on hand sorting of the paper grades and have NIR sorters on the plastic fractions. Due to the pressure from the infeed, where material is becoming increasingly contaminated and requires additional staff to clean up the recyclates, many of these facilities are retro fitting on their paper lines as they develop.
Typically, the mega MRF will have 24-27 pickers per shift on all materials with a wage of around $12 per hour (including costs), which is similar to the UK at £9.60ph. This results in a staffing cost of around $15 or £12 per tonne.
If these mega MRFs were built with more automation such as additional NIRs on the paper grades, the CAPEX costs would increase to £12-15m and OPEX reduce to a more competitive £5-£5.50 per tonne.
With comingled collections becoming more popular, the USA faces the same issues as the UK, where the quality of the infeed materials is decreasing, so labour costs are increasing to produce the same quality material. So we predict that automation will be a key development in the upgrading of existing and future MRFs.
What becomes clear is that by duplicating what you have already is not always automatically better or more cost-effective.
A number of key factors have to be considered before an upgrade, upscale or new-build decision is being made. Some of these, like labour availability and cost or infeed quality, can be outside of the control of the MRF operator or designer. Nevertheless, they cannot be ignored.
Market research and detailed plant design planning, with expert advice, are a good starting point to make an MRF as profitable as possible.