IMANI alert: The AMERI Power deal just got murkier

We have seen a publication purporting to be a comprehensive response from Dr. Kwabena Donkor, (DKD) the former Power Minister, regarding our relentless campaign to expose how extremely horrible the AMERI deal. (see publication http://starrfmonline.com/2017/04/24/kwabena-donkor-responds-to-bright-simmons-ameri-non-facts/ )

We expected to see some rational answers finally. We were sorely disappointed.

Where the argument and arithmetic were lucid enough for us to follow, the conclusions were simply shocking.

Firstly, we will ignore all the ad hominem attacks trying to paint this as another NPP – NDC political slugfest. Such is a typical game of many politicians in this country, with which we are all too familiar. We won’t fall for the bait and get into a terrain that many a politician feel so comfortable operating in.

We have only one agenda: the thorough and precise stripping bare of the AMERI deal so that everyone can appreciate this most brazen of a deal.

1. The LM2500+ vs TM2500+ Issue

Our argument is that the TM2500+ is MORE THAN JUST a gas turbine. It is a gas turbine PLUS various accessories which together make up a POWER PLANT IN A BOX.

Dr. Kwabena Donkor (DKD) goes to the GE website and reproduces a statement from GE that the TM2500+ is a “Gas Turbine and Power Plant” and still proceeds to argue that we are “re-engineering the name”.

All we want the public to note is that the TM2500+ is the COMBINATION of the LM2500 (which is the actual turbine inside the TM2500+) and various accessories to make a virtually complete, compact, power plant solution. The LM2500 costs roughly $9.5 million, while the other accessories add an additional $12.5 million of components needed to create a portable power plant (see: http://www.gas-turbines.com/trader/kwprice.htm). These are “retail level” prices. When bought in bulk or directly from the manufacturer, the price is likely to come down.

As evidence for this point, we have attached below an extract from an offering made by Derwick, a Venezuela firm, to PDVSA, the Venezuelan national oil company, which has been the subject of intense inquiries by US agencies. In that deal the TM2500 was offered without its accessories for $11.5 million. This price obviously included the markup.



The key point being made here is that after paying nearly $22 million per plant, one does not need to spend a huge chunk of resources on top to start producing power.

1. Why So Much Inconsistency?
If the TM2500+ plants were indeed presented to Parliament as costing $22 million each, one has to wonder about the motives of various former Government functionaries who consistently produced different figures in various media engagements in recent weeks.

In the Newsfile edition of April 15th, 2017, Dr. Ayine was categorical in his claim that the ten TM2500+ plants costs $240 million.

In the April 8th, 2017 edition of the same program, Mr. Jinapor, the former Deputy Power Minister, routinely denied that the power plants cost $22 million per unit and insisted that the set of ten costs $360 million.

This is an extract from a statement issued by the same Kwabena Donkor on December 14th, 2015, in which he said as follows:

“It must also be explained that the quoted price of $220 million in the Norwegian story for outright purchase of similar turbines is exclusive of all other costs such as auxiliaries, balance of plant, civil works, sub-station, installation of equipment, cost of financing, operation and maintenance etc (See: https://yen.com.gh/31470-power-minister-denies-duped-norwegian-fraudster.html)

In the statement above, he mentions that the $220 million is “exclusive of ALL OTHER COSTS including AUXILIARIES”. This of course is completely false because the TM2500+, when one buys the high-end package, does come with a full set of auxiliaries detailed in the product brochure here: http://www.aprenergy.com/sites/default/files/tm2500_mobile-gas-turbine.pdf. In fact, the denial of the incorporation of the auxiliaries has been one of the three main mechanisms through which this whole horrible deal was perpetrated.

But let’s save this point for later. What is important at this stage is that we now have admission that the cost of the platforms in total came to $220 million.

We insist that this price was inflated as it implies a cost per MW of $880,000. We have reviewed several TM2500+ deals that ranged from $450,000 per MW (a benchmark often used for estimating the capital costs of aeroderivative power plants in the energy plant costing literature) to $850,000. We note a deal in Egypt where the per MW cost was $500,000.

But considering that there are far graver defects in the whole enterprise, let’s not dwell too much on this point. Let’s go with the $220 million as the actual cost of the set of ten TM2500+ power plants.

2. TM2500 & LM2500

Just so that no one is confused on this point: it is DKD who appears not to know his facts here. The LM2500 is the power generating turbine that has been packaged with other accessories to create the TM2500+, a power plant in a box. This fact is amply documented by GE itself, for example here: http://site.ge-energy.com/corporate/powerxpand/downloads/GEA18664_TM250_r2%20new.pdf
It is also a fact that buying the turbine system alone without the major auxiliaries that DKD claims does not come with the package would be considerably cheaper than $22 million. In fact, many vendors offer that stripped-down package offer it at less than $9 million (example: see http://cfaspower.com/GTG_2048BH_LM2500DF.doc).

3. Integrating a single TM2500+ Plant vs Integrating a System

DKD claims that integrating a single TM2500+ into a national grid is substantially different from integrating a set of several. This point makes no sense. Each TM2500+ system comes with an onboard electrical system to enable integration into a national grid provided the skids, generating stepup transformers and inter-ties are supplied. Integrating more than one merely requires more labour and materials. The time factor doesn’t really change unless one is unwilling to provide the right amount or level of manpower.

4. Time for Integrating a TM2500+ into the Grid

The claim that the TM2500+ can be integrated into the grid for power production within three days did not come from IMANI. This is actually a well-established fact in the industry (see, for example, http://www.industrialmarinepower.com/-3-2011-aeroderivative-gas-turbines-tm2500-mobile-gas-turbine-generator.html). The whole argument is diversionary piffle.

5. Balance of Plant & Operations & Maintenance

Neither IMANI nor Bright Simons has sought to misrepresent the claims of DKD. The claims are usually self-incriminatory enough. DKD has always insisted that the Balance of Plant, installation and Operations & Maintenance come to $118 million thereabouts, WHICH IS EXACTLY THE CONTENTION WE DISPUTE.

In all his public commentary, he has never split this global figure among these different categories of cost. Finally, he has done that and made the job of exposing this massive extortion perpetrated on Ghana all the easier.

6. The Great Cost Inflator! Balance of Plant

According to DKD, to install, integrate into the grid, maintain and operate the 10 TM2500+ in order to generate power, the following costs were incurred:
A. Balance of Plant = $27.084 million

B. Operations & Maintenance = $84 million

C. Insurance Premiums = $6 million

The remaining amount of $171 million of the fixed capital cost recovery payments must then have gone to financing. Now, let us take you through the crazy expenses that were racked up in the name of Balance of Plant. Despite the very loose specifications, our knowledge of the TM2500+ still enabled us to conduct rather rigorous price reviews. Bear in mind also that labour costs have been accounted separately in the bill of quantities supplied by AMERI/DKD.

Engineering and Home Support was billed at $1.5 million for a turnkey platform despite the fact that the manufacturer has produced every schematic necessary for assembly and interconnection with all manner of transmission systems, and has supplied manuals to boot. This inexplicable cost sets the tone for the rest of the rather shameful costing schedule for the AMERI project.

A. 5000 feet of 3C 500 MCM cable was bought at $375,000, clearly inflated by as much as 50% of the open market retail price (https://www.anixter.com/en_us/products/Control-and-Power-Cable/p/7D-5003A)

B. It is important to remember that the interconnection and termination cabling, which is usually a very expensive component of overall cabling, was performed per the contract by Government of Ghana (GoG) at GoG cost. The relevant portion of the Annex E(7) to the BOOT-PPA agreement is produced below.



A. The two 180 MVA transformers capable of stepping up generated power for supply to the GOG interconnect system at 161 KV (and 50 HZ frequency) were billed to Ghana at an outrageous price of $4 million! Whether ‘shell’ or ‘core’ this price is simply unacceptable. Our extensive market survey of prices for the specifications provided came up with an average of $1.1 million, suggesting price inflation of 90% plus.

Transformers are commodity items, now available on unsophisticated marketplaces such as Alibaba (see, for instance, https://www.zauba.com/export-MVA+GENERATOR+TRANSFORMER-hs-code.html), where with a simple email one can easily secure quotes of less than a million dollars for platforms of the same spec as what we required. Even high-end sellers like Compton Greaves and Hitachi nowadays offer one-stop project-based quotations for CIF and FOB pricing. It is rather frightening to note that DKD and his crew are boldly dangling these prices in the face of the public.

B. Ghana was billed a mindboggling total of $9 million for 5- and 6-inch bus ducts, together with their breakers and relay systems that connect the generating units, the step-up transformer and associated equipment!
Using the rather comprehensive frameworks provided by Pankaj Rajput (http://www.neoasis.in/wp-content/uploads/2014/10/Project-Report-Busduct-Design-Cost-estimation.pdf) and the Worley Parsons costing model for the New York Electric & Gas Corp (NySeg) CAES project, we have worked with multiple configurations (guided by the MCM cable length data) and no matter what we do it has been impossible to find out how they could have spent more than $3 million on bus ducts!

What DKD and AMERI is telling us is that they spent nearly $10 million on wiring and cabling. Yes, these are specialty wires and cables, but for a project of this scale, the notion is preposterous!

Since F.V. Smith’s seminal work on the subject in the fifties, auxiliary wiring and cabling has always been estimated and validated by multiple studies as contributing between 1% to 2% of total capital costs of power plants (overnight levelled). To see a contribution of nearly 10% in these AMERI project documents is stupefying to say the least.

[In all this, it is absolutely important to note that the TM2500+ has its own pre-packaged electrical balance of plant that contains all the components that have been costed at over $27 million in this murky deal, provided the customer goes directly to GE, as was the case in the Libyan deployment not too long ago (see http://www.us-lba.org/latest-news/ge-showcases-ebop-capabilities-providing-portable-backup-power-libya-just-six-weeks).]

C. Given that the GOG is tasked with providing treated fuel that meets the exact specifications of the plant, the gas conditioning plant at $500,000 is redundant and needs to be justified.

D. Shipping on CIF incoterms basis has been pegged at – wait for it - $9.434 million! Note that in this case customs duties were waived, so we are concerned strictly only with freight and insurance.

The total weight of a TM2500+ is less than the Gross Vehicle Weight (GVW) of an articulated DAF truck with trailers. For each of these platforms to ship on CIF basis at nearly $1 million completely beats the imagination. It is simply not possible to fathom how the CIF minus customs rate of ten articulated trucks can hit $9 million nor is it fathomable for the TM2500+ plants. Using an extreme range of TEU values, insurance quotes, and after checking with multiple freight forwarders, we find that it is literally impossible to justify more than a $1.5 million cost for this item.

Whilst the Balance of Plant costings are overwhelming in the degree and extremeness of their inflation, they are nothing compared with the O&M costs that were used to compute the final bill.

1. The Great Dupe! O&M Costs

The jargon being used to obscure the operations and maintenance costs in the AMERI deal is “Long-Term Service Agreement”. According to the contract, AMERI is responsible for maintaining and operating the plant. For this work, they have been given a cool $84 million in the cost buildup. In essence, therefore, more than 25% of direct project costs are to go to maintenance and operations.

In previous essays on this topic, we have provided extensive evidence from a wide range of sources that show that the costs of maintaining and operating an aeroderivative power plant rarely exceeds $20/kw/year, which in this case would mean less than $5 million a year and a total of $25 million in 5 years. And ABSOLUTELY and certainly NOT $84 million.

Below is an extract from a quotation sent from ProEnergy Services, a well-known plant maintenance and operations services company, to Derwick that comes to $5.7 million. The disclosure comes from the well known investigation of US federal authorities into the Derwick – PDVSA deal in Venezuela.



Here is another one by Gillian Charles for Northwest Power & Conservation Council that comes to $17/kw/year - https://www.nwcouncil.org/media/6940212/Draft7pSCCT.pdf. There are literally thousands of models worldwide that all converge within this reasonable bound.

And, as we have repeatedly stated, using the global best practice costing libraries of Thermoflow’s GT Pro and Gas Turbine World Handbook all corroborate this $5 million per year average cost (benchmarked to the output of the Ameri Plant Complex). So by which magical means did AMERI and DKD succeed in arriving at $84 million over the term of the agreement?

As far as the insurance cost of $6 million is concerned, we will simply refer our readers to this well written analysis that pegs typical insurance costs at 0.6% of the overnight costs of power plants (http://www.brattle.com/system/publications/pdfs/000/005/010/original/Cost_of_New_Entry_Estimates_for_Combustion_Turbine_and_Combined_Cycle_Plants_in_PJM.pdf?1400252453). This is a benchmark we have seen corroborated widely, and which suggests a more reasonable insurance bill of less than $1.5 million.

It should be evident now where the bulk of the $118 million in so-called Balance of Plant and O&M costs went. A whopping $84 million plus of the money is pure padding. The remaining “hole” of $66 million that can’t be properly accounted for is to be found in the sweetheart financing component of $171 million. Before we get to that though let’s address the remaining points of misinformation in DKD’s rebuttal.

1. Dr. Donkor’s Fuzzy Arithmetic

In point 4 of his essay, DKD insists that there is an annual ceiling for the variable charge that is pegged at: $9,263,700. And that when this is added to the aggregate BOOT charges, the sum should come to $510 million. There is no miracle of arithmetic that will make this possible.

In Annex G to the Agreement bulldozed through Parliament, the monthly fixed charge is: $8.5 million. The aggregate of that over 5 years is $510 million. One does not have to add anything else to obtain the $510 million figure. It emerges as the inevitable outcome of the calculation.

In the same Annex G, the annual, maximum, variable cost is pegged at $16.6 million. Where, Dr. Donkor, are you pulling this $9.2 million maximum cap from? Are you saying that the amount we really ought to be paying is half what is in the contract? How come? How strange! But even so, the total cost to Ghana (fixed plus variable cost) will still come to a hefty $556 million-plus over five years, which is not small change. That whole section of DKD’s statement should be disregarded for incoherence.

2. The Great Fuel Conundrum

The third mechanism (in addition to the Balance of Plant and O&M mechanisms) through which the monstrously shameful AMERI deal was perpetrated was via the deliberate confusion over fuel accounting.

DKD cannot distinguish between the phenomenon of fuel being delivered by VRA but being paid for by the recipient power plant and then recovered through the approved bulk generation charge billed by that power plant to the distribution company (ECG, in our case) and another phenomenon where a power plant does not pay for the fuel at all and therefore does not add it to their charge to begin with.

To obtain the true tariff rate of the company receiving free fuel, paying no income tax, and freed from the costs of land leasing and regulatory compliance, one MUST necessarily compute those cost factors and add them to the tariff before proceeding to compare the tariff to others within the same class as that power producer.

This point is so elementary that we shall not bother to overflog it. When the full costs borne by Ghana on a net basis of the power produced by AMERI is computed, it does work out as:

A. The power producer generating the most expensive power on capacity charge basis.

B. The power producer generating the most expensive power on a levelised cost of energy basis &

C. The power producer with the highest capital costs per MW currently operating in Ghana today.
We note that DKD would like to hide behind some of the equally egregious deals signed under the previous administration, but to the extent that those deals haven’t fully materialised in terms of generated power yet, AMERI remains the most extortionate scheme foisted on us.

3. Hiding Behind PwC

DKD has finally released what he probably believes is his trump card, the mysterious “value for money audit” conducted by PwC, AFTER the deal with Ameri had ALREADY been struck, and Ghana saddled with a burden it cannot escape.

The truth is that PwC is itself very modest in the claims it makes in that document. It did not use many of the specialised databases available for the kind of calculations it needed to perform to come to the conclusions necessary to provide the kind of vindication DKD thinks they have provided, but which on a closer reading of the report were not actually provided at all.

A. Firstly, in the “preamble” to the document, PwC disavows the authenticity of its own sources. It makes it clear that it has taken no extra step to validate the sources used.

B. PwC clearly advises against reliance on the representations for any other purposes other than as confidential advice to the then President, who commissioned them. They clearly did not anticipate the public discussion of the claims made therein.

C. In a future document, we may express in better detail our full appreciation of the claims in the PwC document, but suffice it to say that in point 2 of the summary of key points, PwC indicates clearly that when ranked with the comparable power plants that it surveyed, the capacity tariff for AMERI is the highest.

D. The approach of levelising over twenty years when full capital cost recovery is guaranteed after year 5 ignores the discontinuity in the series and cannot be fully relied upon for comparison purposes.

E. PwC’s decision to compare AMERI, an aeroderivative fast-track solution to combined cycle plants and the like, and in particular to choose three companies that are in various stages of incompleteness (Jacobsen, Amandi and Cenpower), reliant on no-recourse financing etc, in order to compute per KW costs when we have in this country completed and operating plants that can be better ‘normalised’ for comparative purposes is strange beyond measure, but it does not change the fact that as far as currently operating plants are concerned AMERI is the most expensive.

F. PwC seems to have been poorly briefed. They even appeared to believe that the civil works component had been billed to AMERI when in fact it had been billed to GoG.

G. The PwC analysis does assume exaggerated fixed O&M costs for AMERI that we have debunked in the early pages of this document.

H. In view of all the above, the IRR analysis may be discarded. We will repeat our claim that taking all the freebies enjoyed by PwC into account, the IRR for the deal as it stands inches close to 40% and is at least 37%, which is twice the typical rates in the industry. Simply put, such a deal would not have passed a proper PURC review. More importantly, this is where the remaining $65 million of the “missing” $150 million is hidden!
I. On page 22 of the report, PwC details a number of key constraints that in our view undermined the analysis. Notable was the absence of critical data from energy sector stakeholders.

4. The “No Money” Sleight of Hand

DKD continues to repeat his mantra of “we went for this shady deal because we had no money”. That argument is root, stem and branch, BOGUS!

We had money enough to commit to payments of close to $10 million a month. We had enough credit to secure a $51 million standby, revolving, letter of credit. We had money enough to accept to pay for the fuel costs of the plant and bear numerous escalation risks. Frankly we had so much money that we opted to forfeit financing arrangements available from one of the world’s wealthiest companies and sign a deal with a company with no discernible track record and intermediated by not one or two but four middlemen!

Where in the world can a borrower/customer willing to put cashflow of such amounts behind a deal, and to provide such credit enhancements claim poverty as an excuse for not going for clearly cheaper options?

The alternative financing calculations provided for rental and outright purchase are simply too ridiculous to warrant further examination. GE has successfully closed hundreds of TM2500+ deals around the world in recent years, either directly or through its preferred partner, APR. Except in the sorry case of the Derwick – PDVSA saga (where prosecutions in the US are pending for major actors in that deal) were strange middlemen used to any extent remotely like this one for the securing of TM2500+ plants, and even there we did not see this degree of intermediation.

There are countless TM2500+ deals where the cost to the final customer came up much lower than we have seen in all the so-called options constructed by DKD and his merry crew of operatives. As we have shown repeatedly, the cost per MW in capital terms and the costs in kwh terms have been lower in all of these comparative deals than what we have been told were on offer in the purported options constructed by DKD.

We have cited the cases of Libya, Egypt, Indonesia, Burkina Faso and Benin in several essays on this topic and need not repeat ourselves here.

We shall not relent in our campaign to promote the renegotiation of the AMERI deal. DKD and his supporters can keep defying the patriotic will of the people of Ghana. As patriots we can only do so much. The ball is in the court of the Government of the day. But in the ultimate analysis, the people of Ghana shall prevail.