Responses by Matshela Koko to questions from News24

Responses by Mr Matshela Koko to questions from News24

Good day Mr Koko,

I hope my email finds you well.

As you are no doubt aware, News24 has for some time now published articles emanating from The Eskom Files.

Among the documents are data and documents that draw attention to the increase and decrease in EAF for the generation fleet between 2016/2017 and 2017/2018.

The increase and decrease are captured in the graph below. You will note slight percentage point differences between this graph and the official statistics of Eskom, however the differences are irrelevant for the purposes of my query to you – which relates to the approximate 7% increase in EAF in 2016/2017 which prevailed in 2017/2018, only to decline again by approximately 7% in 2018/2019.

There has been some discussion inside Eskom over this increase and decrease in the EAF, and internal documents were prepared setting out possible scenarios for this.

I am hoping that you can assist News24 by responding to some questions over this issue and matters arising from your public statements regarding the performance of Eskom and what you have previously called “the Molefe and Koko effect” – with reference to the fact that during the same EAF increase period, there was no load shedding.

1. In your view, what contributed to the increase in EAF for the period 2016/17 and 2017/18?

The Practice note dated 1 February 2016 was the single biggest contributor to the increase in EAF for the period 2016/17 and 2017/18?1

1 I sent you this practice note via WhatsApp. You will also find it on my twitter timeline.

It brought about operational discipline and high-performance culture.

Our slogan was, “The Power Station Manager is the King”.

This practice note made the Power Station Manager the King.

The Power station Managers were given the full authority for their power stations. This included hiring and firing. It also included prioritization and optimization of operations and maintenance activities in the manner that reduced incidents of capability loss.

We made sure that Power Station Managers have the money they needed to execute maintenance. That no critical plant stood idle and unmaintained because there was no money.

In return we held the Power Station Managers accountable. We also rewarded them unashamedly

The basis of the Practice Note was the following formular

Poor Performance (PP) + Good Excuses (GE) # Acceptable Behavior (AB)

Poor performance plus good excuses were considered not acceptable and when not cured poor performance was a dismissible conduct.

Loadshedding was an example of poor performance

Age of plant, coal qualities and unavailability of operating reserves were examples of good excuses

Power Station Managers were protected for honest mistakes and having grown up through the ranks I knew what an honest mistake looked like.

2. In your view, what contributed to the decrease in EAF in 2018/2019?

I have no personal knowledge of what happened at Eskom in 2018/2019. I left Eskom in February 2018 after an illegal instruction by Mr. Ramaphosa which was directed to the new board of Eskom that was yet to meet to dismiss me.

The Eskom spokesperson is better placed to respond to this question. Ek was nie daar nie.

3. Do you believe that the so-called ‘maintenance festival’ in 2015 contributed to the EAF increase? What role did you play in this maintenance prior to your appointment as head of generation and later as GCEO?

I firmly believe that the maintenance festival in 2015/2016 contributed immensely to the EAF increase. The plant capability loss factor of 12.99% was the highest in the history of Eskom.

It has also been confirmed by the regulator, NERSA, in its Multi Year Price Determination Report for the period 2019 to 2022, at page 53. It states,

“It must be noted that in the years that Eskom focused on aggressive maintenance, plant availability improvements were realised. This is evident from the improved performance seen in FY2016/17 and FY2017/18”

The delegation of authority for scheduling, planning, and executing planned philosophy outages was granted in writing by the board of Eskom to the Chief Technology Officer of Eskom.

The centralized maintenance services of Eskom reported to the Chief Technology Officer.

I was the Chief Technology Officer (Group Executive: Technology) of Eskom from 1 April 2014 until 23 October 2015. I then assumed the role of Group Executive (Generation and Technology) there after until 1 December 2016 when I was appointed the Interim Chief Executive.

Consequently, the scheduling, planning, and execution of preventative philosophy maintenance happened under my direct control from 1 April 2014 until I left Eskom in February 2018

4. What in your view, has contributed to the dramatic increase in unplanned capability loss factor starting in 2018/2019?

As in Question 2 I have no personal knowledge of what happened at Eskom in 2018/2019. Ek was nie daar nie.

5. Please see the below graph. What in your view, contributed to the decrease in utilization factor at the same time as an increase in availability factor, and vice versa? (The graph shows key indicators for coal fleet only)

The above graph is a bastardized version of the graph that was presented by Andre de Ryter on 15 March 20201 during the state of the system adress. I present the actual graph below. It speaks for itself. I agree fully with the key insights on the slide as presented by Andre de Ryter.

6. Do you believe that your Card System for power station managers contributed to circumstances where same managers would run units long after a stoppage was needed, ie, to avoid disciplinary steps in the form of a card?

This is very disrespectful to the power station managers I know, and I have worked with. The the power station managers I have worked with are very competent and are loyal to Eskom. They would never do anything stupid. They are not suicidal. They are not thieves either.

The engineering definition of running “units long after a stoppage was needed” is called an operating excursion. An operating excursion means a generating unit was operated outside of its operating and technical specifications.

The Power Station Managers I worked with would refuse to operate generate units outside of their operating and technical specifications. This is against the PFMA. They are Guardians of these generating units. They have the duty of care.

In terms of the practice note of 1 February 2016 which is colloquially referred to as a ‘card system’, operating a unit outside of its operating and technical specifications is a misconduct.

A Power Station Manager has no authority to contravene Eskom’s specifications. He would be cautioned with a yellow card for that.

7. Do you deny that your Card System was unconstitutional?

It is either you have not familiarized yourself with the practice note or you failed to seek legal advice after reading it.

The practice note must be read together with the Eskom’s Disciplinary Code (Unique ID 32-1132) and Eskom Disciplinary Procedure (Unique ID 32-1113). In the event of any conflict between the practice note and the Disciplinary Code and Procedure, the Disciplinary Procedure receives preference.

The practice note speaks for itself, and the rules of fairness are entrenched in the procedure.

The practice note is procedural, lawful, and constitutional.

8. Do you deny that the EAF was artificially manipulated during your tenure as GCEO through a practice of not declaring partial load losses on units? The EAF could also have been kept higher if unplanned maintenance was not done. Do you wish to comment? Did you instruct power station managers not to declare partial load losses? Did you believe alternatively, your card system may have made power station managers afraid to report partial load losses and/or trips?

The statement,

“The EAF could also have been kept higher if unplanned maintenance was not done” is illogical if not irrational.

I get the impression that somebody wrote it for you, and you just don’t realise that there is no logic in the statement.

There is no factual basis to allege that “the EAF was artificially manipulated”.

I have not seen any evidence to back up this allegation. As a matter of fact, the allegation is preposterous. It is meant to divert attention from the poor operational performance that Eskom finds itself in today. It is an attempt to undermine the best operational improvements of 2016, 2017 and 2018 under my leadership. Fact is that the Eskom leadership is failing today and news24 must stop looking for scapegoats to cover up Andre de Ryter’s dismal performance.

I am responding to your questions in the middle of loadshedding and it can only get worse.

Eskom's reporting criteria is available on Eskom's website, at

www.eskom.co.za/OurCompany/SustainableDevelopment/Pages/Sustainable_Development.aspx.

SizweNtsalubaGobodo Grant Thornton Inc provided a reasonable assurance opinion on the KPIs, marked with “RA” in the statistical tables of the integrated report2. EAF is one of them and it is not qualified for the last 10 years. 2 Integrated report 31 March 2021 pp 128-135

To point out as you did that that “no auditor in the world would be able to pick up from the reported information whether partial losses were appropriately booked as and when required, as those auditors were not on the ground” is at best lack of experience on your side and at worst you are doing the bidding for Andre de Ryter and you are coming across as extremely irrational.

SizweNtsalubaGobodo Grant Thornton Inc should not be giving “reasonable assurance opinions” when they are not able to do so. You are not really sounding proper.

Loadshedding and OCGTs

9. You have previously claimed that you stopped loadshedding in August 2015. Can you please provide more clarity on this? How did Eskom meet the demand of its customers during your tenure?

The amount of unserved energy from 8 August 2015 until June 2017 was zero. This is common course and there is nothing more to explain.

This is supported by the CSIR loadshedding report of 2021

10. I wish to point you to the following graph, which is TW/h sent out by Eskom’s coal fleet and all generation units. This shows that there was no significant increase in electricity generated by Eskom units during the same EAF increase and decrease period. How in your view, is this explained, particularly in the context of no-load shedding during this time period?

Why would you generate more TWh when the demand curve is flat? It does not make sense.

What you will do when the EAF is increasing because of improved maintenance, and when the demand curve is dropping is to reduce the generation load factor.

The generation load factor for 2015, 2016, 2017 and 2018 was 61.5%, 58.8, 57.9 and 55.9 respectively.

This makes perfect sense.

The demand is flat and the EAF is increasing. The only rational response is to reduce both the diesel consumption and the plant load factor subject to meeting the demand.

What happened after 2018 irrational, and I don’t want to comment on it. Ek was nie daar nie. Mr Mantshantsha has a job cut out nicely for him.

11. Do you agree, that loadshedding was averted during your tenure not through some intervention by yourself or Mr Molefe, but rather because of the increase in maintenance done in 2015, which started prior to you being appointed head of generation or acting GCEO?

As a Chief Technology Officer at the time, I was directly responsible for maintenance in 2015 and before that. See bullet point 3 above.

The board of Eskom had directly delegated to me the responsibility to schedule, plan, and execute preventative philosophy maintenance.

I am happy that you acknowledge that under my stewardship in 2015, maintenance execution was a success and for that I thank you.

12. It is well known that during your tenure, expenditure on Eskom-owned OCGT’s was slashed dramatically. Yet, there was no load shedding and apparently, no dramatic increase in how hard the coal generation fleet was being run. Could you clarify this?

That’s what must happen. I am happy you recognise that. Andre de Ryter has no excuse. The demand is dropping by 2% per annum.

He must be able to keep the lights on with EAF of at least 75%, do all the necessary maintenance at a PCLF of 10% and only burn a maximum of 10 ML of diesel and kerosene per annum.

Anything less than that is incompetence

13. In the light of the above, is it fair to say that the onboarding of Ingula capacity as well as capacity from Avon and Dedisa OCGTs and other IPPs, made up the shortfall?

The new generating capacity will always assist in meeting the demand. It will enable more PCLF, and less diesel spend for a flat demand curve.

What you deliberately ignore and that’s in defence of Andre de Ryters is that there was no Ingula, Avon and Dedisa OCGTs in August 2015 when we stopped loadshedding. We stopped loadshedding without Ingula, Avon, Didisa, Kusile and with only Medupi unit 6 in commercial operation.

14. In your claims over expensive OCGTs being utilised less during your tenure, why have you failed to mention the billions of rands paid for the use of Avon and Dedisa?

I am offended by this statement to make.

The diesel and kerosene usage for OCGTs in Ml is in page 131 of 31 March 2021 integrated report for a 10-year window.

Avon and Dedisa are not included, and the auditors are happy with that.

They can’t be included because they are IPPs. You will find their costs under the primary energy costs. This is so for all IPPs including REIPP. It sickens me that you purposefully ignored this point. I don’t think it innocent.

Follow Up Questions for Mr Matshela Koko - 19 October 2021

Good day,

Further to my previous questions regarding the increase and decrease in EAF, herewith please find follow up questions I hope you will respond to before 6pm today:

In 2016/17 and 17/18, Eskom placed an equivalent 15 million MWh in available cold reserve, and cited in annual reports that this was due to surplus capacity and brought about in part by the “pressure” of IPP purchases on Eskom – ie, the SO didn’t need the units.

It is noteworthy that the largest contributors to the high AVCR levels were not mentioned in annual reports at all.

These are the highest levels of AVCR in the past decade and the 16/17 AVCR of 5.6m MWh represents a 4131% increase over the 14/15 FY. In 17/18, AVCR reached 10.4m MWh, before reducing to 755 000 MWh in 18/19.

Additionally, 5.5m MWh of OMCR (Opportunity Maintenance Cold Reserve) was undertaken over these two years.

In some cases, the MWh OMCR registered for certain power stations in 17/18 actually exceed the AVCR hours.

It is noteworthy that none of this was ever reported in annual reports, and effectively the country was left in the dark about this practice and MWh deemed losses, as per the Plant Reporting Standard (various revisions), are not included in planned or unplanned losses.

For your ease of reference, the power stations with the highest contributions toward the AVCR figures but which were not mentioned in annual reports are; Tutuka, Matla, Majuba and Kriel. It is noteworthy that as of 2020, Tutuka, Majuba and Kriel are among the highest contributors to unplanned losses.

1. Did you issue directives, as this was under your leadership after all, for the units to be placed on cold reserve in an effort to decrease unplanned losses in these same years, thereby creating EAF levels 2% and 3% higher than the prevailing reality?

Brian Dames, Thava Govender, Mongezi Ntsokolo, Willy Majola, Bheki Nxumalo, Rhulani Mathebula and Phillip Dukashe in their capacity as Group Executive - Generation were not delegated to issue directives for the units to be placed on cold reserve. Generation Group was under their leadership after all.

I inherited no special powers. I worked with the same delegation of authority.

The outage approval process was introduced in Generation Group in 2013 when Brian Dames was the GCE and Thava Govender was the Group Executive – Generation. It was documented in a Generation Standard – GGL-36-48 Rev 12. This standard set up the Short-Term Energy Review Forum (STERF).

The requests for all opportunity maintenance events are made to the Generation Dispatcher and the System Operator; and are ratified at the weekly STERF meeting. STERF at all material times acted within its terms of reference.

2. Alternatively, were you aware of this practice?

There was no such a practice in my time. That’s why Eskom is not making this allegation and it will not.

The overall effect of this on the generation EAF is a one-year increase (followed by the increase in maintenance previously discussed with you) and a steady decline thereafter, as opposed to the official two years of stable, increase EAF performance.

I stand by previous responses to you. I repeat them here.

“SizweNtsalubaGobodo Grant Thornton Inc provided a reasonable assurance opinion on the KPIs, marked with “RA” in the statistical tables of the integrated report. EAF is one of them and it is not qualified for the last 10 years.

To point out as you did that that “no auditor in the world would be able to pick up from the reported information whether partial losses were appropriately booked as and when required, as those auditors were not on the ground” is at best lack of experience on your side and at worst you are doing the bidding for Andre de Ryter and you are coming across as extremely irrational and comical.

SizweNtsalubaGobodo Grant Thornton Inc should not be giving “reasonable assurance opinions” when they are not able to do so. You are not really sounding proper”.

3. Do you deny that units placed under cold reserve in 16/17 and 17/18 were in fact, units that broke down and should have been declared as UCLF?

I confirm that cold reserve conditions were complied with for units that were placed under cold reserve in 16/17 and 17/18. I say this because SizweNtsalubaGobodo Grant Thornton Inc provided a “reasonable assurance opinions” on the EAF numbers.

During cold reserve period the unit is available to operate but it is not electrically connected to the transmission system for capacity or economic reasons. Stations require a minimum of 7 days for the System Operator to grant cold reserve status. Once cold reserve is granted on a unit, opportunity maintenance gets executed without affecting the call back time. The call back time is determined by the Systems Operator.

4. Previously I asked if you denied that you instructed or were aware whether power station managers did not report all Partial Load Losses for fear of being disciplined under your card system regime. You failed to confirm or deny this.

I stand by my previous answer. It is unambiguous. I repeat it here.

This is very disrespectful to the power station managers that I know, and 1 February 2016hat I have worked with. The power station managers I have worked with are very competent persons and they are loyal to Eskom. They would never do anything stupid. They are not suicidal. They are not thieves either.

The engineering definition of running “units long after a stoppage was needed” is called an operating excursion. An operating excursion means a generating unit was operated outside of its operating and technical specifications.

The Power Station Managers I worked with would refuse to operate generate units outside of their operating and technical specifications. This is against the PFMA. They are Guardians of these generating units. They have the duty of care.

In terms of the practice note of 1 February 2016 which is colloquially referred to as a ‘card system’, operating a unit outside of its operating and technical specifications is a misconduct.

A Power Station Manager has no authority to contravene Eskom’s specifications. He would be cautioned with a yellow card for that.

5. Can you do so now? Are you now, or were you ever aware, of a practice at power stations where managers would not declare partial load losses, which was done in an effort to maintain lower levels of unplanned losses?

I am not aware that power station managers behaved unethically and irregularly by hiding load losses. The power station managers that I worked with would not do this.

6. There is a variance between official, reported planned maintenance hours as contained in the Generation KPI data, which is higher than the planned maintenance hours contained in the GPSS system by some 20m MWh between 2013 and 2021. The highest variances are during your tenure, ie 5m MWh in 2015/16 and 2m MWh in 16/17. Are you aware of the reason for this?

There is no way that SizweNtsalubaGobodo Grant Thornton Inc would have missed this variance. The Generation KPI data is the GPSS system. I am not aware of parallel systems

Please contact SizweNtsalubaGobodo Grant Thornton Inc on this one.

7. I previously asked you for your views on what contributed to the decline in generation performance starting in 2018. I ask again, but must ask whether you are now, or were ever aware, of any practices at power stations that may have contributed to increased unreliability in subsequent years, including but not limited to running units under partial load losses for extended periods – sometimes a year or more – to prevent disciplinary action being taken against them as a result of your card system?

Ek was nie daar nie. Please refer this question to Phakamane Hadebe and Andre de Ryters. I cant speak for them

8. Do you deny that the performance credit criteria as stipulated in your practice note of 1 February 2016, incentivized power station managers to run units continuously, despite faults, to prevent disciplinary action being taken against them?

The practice note of 1 February 2016 incentivized power station managers to operate the power station within their technical and operating specifications.

They have done very well in this regard

9. In support of the allegation in question 5, I draw your attention to Kriel Unit 5. Starting prior to September 2015, the unit ran on partial load because of a “ID fan vanes at maximum load” fault until the turbine ceased on 5 March 2018 – shortly after you resigned. 25 days of PCLF was booked on the unit between September 2015 and March 2018, before the turbine ceased. (The unit went on GO outage shortly after, for 100 days). If you deny that your card system and the prevailing culture in generation during your tenure did not incentivize power station managers to run units past breaking point, how do you explain events at Kriel, which was not isolated only to unit 5?

I don’t think you understand this question.

You are not saying to me that the furnace pressure control was outside the operating and technical specifications in this case study. For all I care the unit can run forever with a partial load loss until the next outage (GO) opportunity for as long as the furnace pressure is controlled within -2kpa and +3 kpa. For as long I have a stable fire ball and a reasonable assurance that the furnace pressure protection will not operate I will run the unit until the next outage opportunity.

Your point on “units passing break point” in this example is extremely nonsensical

In March 2018, Ek was nie daar.