The Hidden Factory, How Big is Yours?


All businesses have to make a profit, or they will probably cease to exist, some businesses operate with small margins whilst more successful businesses enjoy greater margins. One thing I am sure of though is that all businesses, no matter how profitable, suffer from the ‘cost of poor quality.’

These costs fall into two categories;

  • those that are easily measured and accounted for,
  • and those that are not easy to measure, if indeed, they are measured at all.

Those that are not easy to identify and measure make up the ‘Hidden Factory’.

The Hidden Factory

The ‘Hidden Factory’ is literally in our faces every day,hidden factory but not many of us see it, it hides in many dark corners of all businesses, even the best-managed business can be hampered by the ‘hidden factory’, and worse still, …….. not even realise it.

There is a financial burden for any company, whether service orientated or manufacturing based, due to the cost of poor quality (COPQ). However, no matter what the business orientation, it is astonishing to think that the hidden cost of poor quality could be a much as 20% of a companies’ revenue. Taking into account that the easily measured costs of poor quality make up 5 to 15 % of revenue, then the total encumbrance of the visible and invisible costs could be a staggering 35% of income.

Profit & Loss

I would suggest that many companies are complacent in their measurement and appreciation of the true costs and that a lot of these costs are bundled up into profit and loss nominal accounts, making them hard to recognise and even harder to address. As long as profit margins remain within budgetary constraints and the status-quo is maintained, why would any unsuspecting business owner be anything other than satisfied with an expected performance.

7-wastesImagine though, if these hidden costs could be identified, and rigorous steps taken to remove them, the effect on profits would be breathtaking. Eliminating the ‘hidden factory’ and addressing the traditional costs of poor quality is one of the many specialities of a Lean Practitioner. Skilled in concepts, such as Kaizen, (incremental, continuous improvement), Value Stream Mapping and Kanban, (just in time process control), to name but a few, a Lean Practitioner is fully equipped to positively affect the profitability of any business.

How can it happen?

In part, a ‘silo mentality’ could be instrumental in the development of the ‘hidden factory’, e.g.
how do the finance, commercial and operational departments communicate with each other.When was the last time a senior member of the finance department visited the factory floor and questioned excessive inventory levels, process cycle times, process constraints, etc.  Likewise, do operational managers question their departments’ performance if they are within ‘budget’, or is it the case that the budget, is the performance measure.

The Silo Mentality as defined by the ‘Business Dictionary’ is a mindset present when certain departments or sectors do not wish to share information with others in the same company. This type of mentality will reduce efficiency in the overall operation, reduce morale, and may contribute to the demise of a productive company culture.

Do the commercial department record  lost sales and customer loyalty, do they win sales on price alone? Who do they inform? Who, if anyone, asks the right questions? Are measurements and controls in place? The list goes on.

What to do?

A simple, cost effective method to determine whether or not your business is suffering from the detrimental costs of the ‘hidden factory’ is to benchmark it against another company, (ideally described as best in class or even world class), competing in the same sector or industry as yours. What is it that differentiates their success or performance from your company? If they happen to be a company that actively practises Lean Six Sigma methodology, then maybe they have acted to eradicate their ‘hidden factory’.

All things being equal, there will be reasons for their superiority in the marketplace which will be evident in their stats and KPI’s. Look and learn!

Doing nothing and accepting the norm is not an option, face the ‘hidden factory’ head on and reap extra rewards without having to raise revenue, become Lean and ‘get more for less’ – I do!




Lean, Six Sigma

A Layman’s Insight into Lean Six Sigma



I am writing this article for the benefit of those readers who have not been exposed to Six Sigma or Lean methodologies and aim to explain some of the basic goals and terminologies.

It’s widely believed that Lean Six Sigma is primarily used for the improvement of manufacturing processes, this is not the case, it can be used to significant effect on any process, of any description.

I will give brief descriptions of Six Sigma and Lean, they compliment each other but improve processes in different ways.

Six Sigma

In simple terms the Six Sigma methodology seeks to eradicate all variation in a process, no variation would mean that the process output would remain constant irrespective of inputs. The idea is that a process without variation would necessarily be a process without defects.

In practice, this would be impossible to achieve but accomplishing Six Sigma would come very close to it. In terms of defects per million opportunities (DPMO) a process with a Sigma score of 6 would only have 3.4 defects per million opportunities. In comparison, a process with a yield of 99% would have 10,000 defects per million opportunities.

Practical examples:facts

The above examples show the marked contrast as processes approach Six Sigma, such industries in this category maybe the pharmaceutical or airline industries. Considering a McDonald’s burger is almost identical the world over, perhaps their processes are approaching Six Sigma and their burger production process is variation free. However, in general, manufacturing processes have an average Sigma score of 4 allowing for 6,210 DPMO, or 99.38% efficient. I have measured processes in the waste industry with a Sigma score as little as 1.0 or 30% effective.


Lean improves processes by eradicating ‘waste’ in processes, it recognises 7 inherent waste types, transport, overproduction, over-processing, defects, inventory, waiting and motion. I also like to recognise ‘wasted’ talent as the eight waste.

Consider the process of baling RDF in the waste industry but imagine as I discussed in another post that operations might produce RDF just because the waste is on site and the plant is sat idle.

  1. Transport – all transport is a direct cost and hence waste unless the business process is transport, i.e. ‘furniture removal’ etc.
  2. Overproduction – producing bales that are surplus to requirements, without an immediate off-taker.
  3. Over-processing – making a bale out of RDF and wrapping it up in plastic, without an immediate off-taker.
  4. Defects – Bales returned because of damage, RDF out of specification, etc.
  5. Inventory – The cost of storing bales for any period and the attendant risks, fire, environmental, etc.
  6. Waiting – Any idle time associated with the baling process, machinery, manpower, transport queues, etc.
  7. Motion – the act of moving a bale around the site before shipment.


Imagine the wasted costs, if bales of RDF are not exported off site to an off-taker. Kanban or ‘just in time’ operations would go someway to preventing this situation and are also Lean tools.

Lean is also used to great effect in assessing the ‘value stream’ of a process. By reducing waste, process cycle times can be reduced, and it serves to rid a process of ‘non-value add’ activities i.e. those parts of a process for which a customer is not willing to pay.

Lean Six Sigma

Both methodologies are successful in their right and are used separately to address different process problems. When combined however, I do not believe a process exists that cannot be improved and that the breakthrough improvements in many processes would be breathtaking and far more than reasonable expectation.

Six Sigma, Waste Management

Producing a specified Refuse Derived Fuel (RDF)


It’s easy to produce Refuse Derived Fuel (RDF), you just get some waste, shred it, screen it to remove ‘fines’, extract ferrous and non-ferrous metals, bale it then wrap it.

What’s the spec., who knows, who cares? However, the off-taker, end customer, will care because the specification of this RDF will significantly have an effect in respect of economic, technical and environmental issues.

Typical specification requirements would be:

  • Calorific value (typically a maximum or within a range, economics)
  • Moisture content (maximum, economics)
  • Biogenic content ( usually a minimum for ROCs subsidy, economics)
  • Chlorine content (maximum, causes corrosion, technical)
  • Ash content (maximum, performance issues, technical)
  • Heavy metals content (maximum, emissions, environmental)

In many cases, these specifications are contracted, and failure to meet them could result in rejected loads, financial penalties, contract termination, etc. The importance is obvious, that the process to produce RDF at an agreed specification needs to be controlled, have repeatability (consistent output), have reproducibility (consistent output irrespective of process operator) and would require a robust measuring system.


Six Sigma prescribes that to have a process deliver a constant specified output (without defects), that there must be no variation in the process inputs.

Consider the inputs to an RDF production process:

  • Waste feedstock type and composition
  • Waste feedstock suppliers
  • Production plant equipment settings
  • Production plant equipment wear
  • Production plant equipment cleanliness
  • Plant operatives (experience, training, etc.)
  • Plant operator work patterns
  • Sampling system (BS 15442)
  • Analysis lab process (UKAS accredited)

How on earth can input variation be reduced to zero in an RDF production process? The practical answer is that it can’t. However, Six Sigma methodology can be used to significantly reduce the different causes of the variability to afford RDF producers the comfort that their process can produce RDF to a given specification or conversely that their process cannot provide the required specification.

Ways that Six Sigma could help in this process are too many to discuss in this post, but I will give some examples of the methodology to give interested readers a little insight.

In any process, there are two types of variation, Special Cause and Common Cause. Common cause
variation is difficult to address because it is likely to be inherent in the process, environmental conditions for instance. However, special cause variations can be affected because they are physically input into the process, such as changes to equipment settings, personnel shift changes, feedstock type, etc.


Six Sigma methodology can be employed to measure the effects of these variations on output RDF specification and determine which caused the greatest impact. These variations are termed the critical few and are improved first. These improvements may be simple and easy to remedy; others may be difficult and need more advanced methods such as Design of Experiments (DOE) which is used when variables in one piece of equipment act to effect the variables of another piece of equipment and so on. Analysis of Variation, ANOVA is then used to determine the significance of each variable interaction statistically.

It can be seen that entering into specified RDF contracts should not be taken lightly, especially when commercial departments consider them with little or no input from their operational colleagues. Even if possible, the costs of attaining unrealistic RDF specification could be staggering.

When considering RDF specification, think of RUMBA:

  • Reasonable – is the spec reasonable.
  • Understandable – is the spec clearly defined.
  • Measurable – is the product measurable against the spec.
  • Believable – is the spec achievable.
  • Attainable – is the spec attainable.

In my opinion, to confidently produce a specified RDF, experience alone cannot be relied upon as there are just too many interacting variables to control. Only the application of a methodology such as Six Sigma would stand a chance to produce a long-term controlled, sustainable, cost efficient solution.


Process Analysis, Waste Management

Predicting the Impossible


Two years ago I was involved as a process analyst to help design a MRF plant that would produce a highly specified Refuse Derived Fuel (RDF) as the feedstock for two Advanced Plasma Gasifiers (worlds largest).


As a team, we drew up a list of criteria that would be required from the plant manufacturer; we invited manufacturers to tender for the plant development based on our criteria and the RDF specification.

Some dropped out of the tender process because they did not believe that it was possible to mechanically produce the RDF specification without some form of extra processing. All had one thing in common, though and that was that they would not guarantee the performance of the plant to produce the RDF specification unless the input raw waste specification could be ensured.

It’s impossible to specify a single load of waste let alone a continuous waste stream because the waste would consist of municipal solid waste (MSW), commercial & industrial waste (C&I), trade waste, and transfer station waste. The building of the plant went ahead with fingers crossed.

Something needed to be done

It was at this point before the plant was built that I decided to develop a continuous process model that would determine the probable mass balance of an MRF plant, irrespective of the waste input composition and so predict the composition and specification of the output RDF. My idea was that if I could achieve this, it would be possible to design a plant that would give the desired output by changing the configuration of the plant by knowing what the input material was and not the other way round.

I set about determining the constituent components of the different wastes using industrial sieves, 28mm, 20mm, 14mm, and 10mm, two microwave ovens and a set of scales.


The work took many months to complete. However, the reward was that I had a full set of data of the composition of different wastes, differentiated into 18 categories, plastic film, wood, putrescibles, etc. I also individually determined their moisture content, density, calorific value and biogenic percentage.

Development of the continuous model:

  • Using the waste by mass compositions, I could mix different wastes together as a homogeneous input in the model.
  • Because I had determined the fraction sizes of the different compositions, it allowed me to determine how different screen types and sizes would perform at different parts of the process.
  • Knowing densities would allow prediction of the air separation equipment performance.
  • For any given input rate the model could determine the throughput rate of material at any point in the plant, this could help size or determine the requirement of conveyors, overband magnets, eddy current separators, optical separators etc.
  • Having equipment with a known operational specification, such as optical separators, for instance, the model could easily determine the most efficient and cost-effective recycling commodity targets and where to optimally place the equipment within the process.


The model produced an accurate mass balance i.e. calculating the individual mass of all the outputs for a given input mass of material:

Input waste = Sum of plant outputs – Shrinkage

A commercial perspective can be gained from this as the model would calculate the percentage of RDF, recyclables and other disposables for any given input waste.

Predicted specification for all outputs would include, moisture content, calorific value, biomass by CV, biomass by mass and density. The model data could then be used to size such things as the plant storage bays, baler capacity, wrapper throughput rates, required haulage by size and type, etc.


The Big Day Arrived

Eventually, the plant was built and commissioned by Turmec with the invaluable help of Brian Thornton. At commissioning, 1,600 tonnes of differing waste inputs in different sized batches were run through the plant over a five day period. I supervised all the measurements of the plant output and meticulously accounted for every tonne. Out of interest, the shrinkage was calculated to be 2%.

I’d modelled the plant configuration before commissioning commenced and inputted the amount and types of input waste as the commissioning phase ended.

I finally ran the model in the presence of Brian; it predicted all the plant outputs with a combined accuracy of <1% (actually it was 0.78%); the model predicted RDF output was 52.10% compared to the actual plant measured RDF output of 51.42%.

The RWM Exhibition 2016

I recently attended the RWM Exhibition at the NEC as a visitor, I visited a lot of stands and had some very interesting conversations with various exhibitors about my model.

I have decided to develop the model further, and will need the collaboration of suppliers to gain knowledge of their equipment technical specifications, etc. Interested manufacturers, when questioned at the RWM, were Turmec, Bollegraaf, Nihot, Magnapower, IFE, Impact Air Systems, Menart, NRT, BHS, Castulik, TEMA Process, Bakker Magnetics, ATM, PRM Waste Systems and LHS.

I intend to visit the RWM Exhibition in 2017 but as an exhibitor, hopefully Predicting the Impossible.