Document(s) 21 de 29

Introduction

Nigeria is a late starter; technologically, it is behind South Korea, Brazil, and India, as is reflected in the research literature. Whereas research in the more advanced developing countries deals with innovative changes in the mature industries (steel, capital goods, etc.), the effort in the present study and in most related research in Nigeria is limited to examining the introduction of technology and the mastery of production technology.

Specifically, this study

• reviews the literature on the transfer of technology in, and the development of, the steel industry in Nigeria;
• undertakes a case study of the Ajaokuta Steel Co. Ltd (ASCL), examining the planning (including feasibility studies and other documentation);
• evaluates the contractual relationships between ASCL and the external suppliers of technology to determine the extent to which a continuing reliance on external participation is assumed; and
• examines ASCL's training and other efforts to develop human resources, concentrating especially on the use and maintenance of transferred technologies.

Practically all projects under implementation in sub-Saharan Africa, with the exception of the Ajaokuta plant in Nigeria, have been either abandoned or frozen. Similarly, with the exception of Sicarta, in Mexico, and Acominas, Cosigua, and some other minor projects in Brazil, all steel projects in Latin America have been either frozen or cancelled. In contrast, China, South Korea, and Taiwan have displayed dynamism, have enjoyed healthy project financing, and have installed capacities at low costs, emerging generally strong in steel technology.

The most common cause of delays and cancellation is shortage of funds. The extensive infrastructures normally required to support steel-making projects in developing countries contribute significantly to the total cost, and few countries can make the investments without securing loans. The external debt carried by developing countries is already so great that creditors have been reluctant to provide additional loans, especially as most developing countries have experienced serious difficulties servicing their debts. Also, the nature and source of financing strongly influence project implementation — projects with a high percentage of local equity, as in South Korea, rarely suffer project delay (UNIDO 1986).

Steel development demands a high level of skills, technical and organizational capability, and systematic sophistication (supply of capital goods and technical services). These components are available (partially or wholly) in most of the Asian nations, whereas they must be imported by African countries and paid for in foreign currency. Lack of financing, therefore, is a sign of more fundamental problems.

Because of the large, diverse, and specialized nature of steel-making equipment, investment in just the technology, especially for integrated plants, is thousands of millions of United States dollars (USD). The cost per unit of installed capacity has risen over time (an average 350 USD/t in 1965 and 1700 USD/t in 1980) and varies markedly between regions (UNIDO 1986). The costs are still rising so fast that time overruns are very expensive, and the increase in costs from delays sometimes makes project completion impossible.

The large capital requirements limit ownership in developing countries to government and, usually, joint ventures (Maxwell 1982). Other characteristics of steel making contribute to the profound effects investments in it have on the economy. According to Maxwell, these characteristics are the indivisibility of the steel plant equipment, the long gestation, the irreversible nature and widespread implications of the process chosen, the idiosyncrasies of each plant, and the complexity of the process. For example, because of the indivisibility of the equipment, the designers (initially or during expansion) prefer to build in extra capacity in units like rolling mills. Overall capacity can later be expanded without unduly large investment in new plant units. These add-ons, which are temporarily (and indefinitely) unproductive, can be a dilemma for steel-plant designers.

Long gestation

Projects like the creation of steel plant are a process: a series of interconnecting actions — past, present, and future — not "phased," or discrete, periods of development. Past actions and outcomes largely influence current and future actions and discussions. In other words, the evolution of a project such as the Ajaokuta steel plant in Nigeria is affected as much today by discussions at its conception as it is by, for example, a decision to add a new slab caster. In fact, the decisions made 15 years ago may have compelled the action taken today. Nevertheless, classifying decisions and events into phases, or categories, simplifies analysis and discussion. Gestation, as understood by Maxwell (1982), comprises all the events from preinvestment through to construction and from start-up to operation.

Preinvestment starts when the company that will build, own, and operate the plant is legally constituted. In this period, planning, feasibility studies, and financial negotiations are undertaken. Political sanctions are also obtained.

Construction begins when preinvestment effectively ends, consisting of civil engineering works, procurement and construction of structural and infrastructural facilities, erection and installation of process plants, dry testing, and commissioning. Detailed engineering and fabrication of equipment may predate this phase. Start-up of steel making is taken to mark the end of the construction period, although other sections of the plant may begin operations later.

Start-up, according to Maxwell (1982), lasts "from the beginning...of production...till the achievement of annual output, which corresponds to...nominal production capacity." Some plants (especially in developing countries) remain in the start-up phase indefinitely, without achieving anything near their nominal capacity, because of a myriad of constraints (management problems, inadequate raw materials, power outages, and absence of ancillary facilities). Maxwell subdivided start-up, defining technical start-up as the period from the start of production until staff learn to operate the plant. Implementation is the sum of time involved in construction and start-up.

The literature on Latin American steel plants indicates that a long gestation period for greenfield plants is the rule, rather than the exception. The time involved ranges from 3 to 19 years, with an average of 10–11 years for both greenfield plants and major expansions. The average is 3–4 years each for preinvestment, construction, and start-up.

A look at Indian plants reveals a similar pattern. Bokaro Steel Ltd was incorporated in 1964; commissioning was scheduled for 1971. Actual commissioning was in 1976. Implementation took 2 years. An expansion, scheduled for completion in 1977, was actually completed in 1984. The first stage of the Bhilai steel plant had a modest time overrun of 2 years, but two expansions had an 8 year overrun.

In the case of Ajaokuta, 12 years passed (1971–83) between the incorporation of a formal body to oversee steel development in Nigeria to the commissioning of the finishing mills (wire-rod mill and light-section mill). A rescheduling took place in 1985, and 23 years slipped by (1971–94) without the full integration of the plant. This record compares with the longest gestation documented in recent literature — the expansion of the Acerias Paz del Rio plant (19 years) and the establishment of the Chimbote (18 years) and Somisa plants (16 years).

The reasons for such long gestation periods are relatively few. Preinvestment is delayed by either a shortage of funds for the capital-intensive operation or a political intervention — usually government's planning of the steel industry. The main reasons for long gestation in construction include suspensions of foreign-exchange remittance by the central bank, delays involved in securing financing for equipment from suppliers, equipment-delivery delays, and organizational problems. Start-up seems to be bedeviled by conceptual errors in the design of the overall plant or some of its facilities; weaknesses or defects in equipment fabrication or plant construction; inadequate preparation of the plant's work force; shortages in supply of key raw materials, such as iron ore and coal; shortages in the supply of key services, such as electricity; and overoptimistic demand forecasts.

One of the key problems is building on a greenfield location — virgin lands in developing countries, which need 75% more spending than an industrialized site and 35% more than a site with some infrastructure (UNIDO 1983). This is demonstrated clearly in Maxwell's study (1982, p. 107):

In the Paz del Rio plant, the lack of availability on time of the planned electrified railway to haul ore and coal from nearby mines to the plant led to some shortages in these raw materials during the startup period... while lack of sufficient electricity supply slowed...production in Paz del Rio and Chimbote greenfields plants.

It would have made more sense to have given priority to...iron ore mining at Itakpe in Kwara State and should logically have started five years before the Ajaokuta works were commissioned.

The problems can be grouped not only by phase but by focus:

• finance — shortages of foreign exchange, high indebtedness, and totalor almost total importation of capital goods, technology, and engineering services;
• planning and management — inexperience, low levels of technological and organizational skills, and lack of precedents;
• politics — control, bureaucratic interference, and shifting government policies;
• execution — behaviour of the contractor involved in project implementation;
• infrastructure — available and required;
• design — concept, fabrication, and modifications;
• supply — lack of critical inputs such as raw materials, spare parts and consumables, power supply; and
• personnel — inexperience, particularly of the start-up team.

The seeds of long gestation are planted long before the effects are noticed; current, as well as future, actions and decisions could be enriched by a careful look at decisions taken ex ante. This demands a review of the evolution of steel making in Nigeria.

The evolution of steel making in Nigeria

Feasibility studies were commissioned, and their results were examined. However, the discovery of iron ore deposits at Udi and Agbaja and coal at Enugu, along with the potential for electricity from the Kainji dam project, then approaching completion, made a case for the establishment of an integrated iron and steel complex.

Broadly, the two steel-making processes involve either mini steel facilities or integrated plants. Integrated plants typically use a conventional blast furnace – open-hearth furnace (BF–OHF) or a direct-reduction electric-arc furnace (DR–EAF). Mini steel facilities are characterized by low economies of scale; low investment per tonne of installed capacity; rapid construction and, thus, reduced time and cost overruns; relatively fast and simple buildup of production; low energy consumption, especially for scrap; and negligible associated environmental pollution. The main disadvantages are a restricted product mix; an expensive main energy source (electricity); and vulnerability to price fluctuation and availability of scrap. Integrated plants have an advantage where markets are large, annually producing up to 10 × 10⁶ t of steel in diverse shapes and items; also they can combine different sources of energy. However, they require large capital outlays, often incur cost and time overruns during construction, experience a slow production buildup, require an extensive infrastructure and organization, and pollute the environment.

Between 1961 and 1965, several firms were invited to submit proposals for the construction of an integrated iron and steel complex in Nigeria. These firms, all of Western origin, were of the opinion that available raw materials could not economically support an integrated steel work with the available technology. A proposal, which was seriously examined, was to build a plant using the "strategic Udy process," which was then at the pilot-plant stage, under development in the United States. The government founded the Nigerian Steel Associates, a joint venture with Westinghouse and Koppers. The project did not, however, take off, as the process did not pass commercial-scale testing.

In 1967, a UNIDO study cited Nigeria as a potential market for steel. A team of Soviet steel experts visited Nigeria the same year to conduct a feasibility study for an integrated iron and steel works. Their study recommended the blast furnace – basic oxygen route to iron and steel making. The study admitted that the local ores were of poor quality and recommended further geological surveys. In 1970, a contract was awarded to Technoexport of the Soviet Union to determine the quality and quantity of the deposits of iron ore, coking coal, limestone, dolomite, and refractory clays available in Nigeria for the operation of an iron and steel complex.

In 1971, the Nigerian Steel Development Authority (NSDA) was formed. The authority was responsible for the planning, construction, and operation of steel plants in the country, as well as carrying out the necessary geological surveys and metallurgical research. The authority also examined various processes, including direct reduction with natural gas, even though this process required high-grade iron ore not known to be available in the country. In 1973, NSDA commissioned Tiajpromexport (TPE) to prepare a preliminary project report for the proposed steel plant.

The report recommended that the plant produce 1.3 × 10⁶ t of long (rods, bars, and light structures) and flat products. The plant was to be sited at Ajaokuta and would use iron ore from Itakpe. NSDA accepted the report's recommendation in 1975, with a few modifications, mainly to produce 1.3 × 10⁶ t of long products in the first phase and 2.6 × 10⁶ t (an added 1.3 × 10⁶ t) of flat steel sheets as soon as expansion was feasible.

In the same year, TPE was commissioned to prepare a detailed project report based on the earlier recommendations. This report recommended an integrated steel works, the first phase of which would include the following:

• a plant with two batteries of 49 coke ovens each (9 × 10⁵ t/year total capacity);
• a plant to process the byproducts of the coke ovens and produce dehydrated tar, ammonium sulfate, etc;
• a plant with two sintering machines (2.64 × 10⁶ t/year total capacity);
• a plant with a blast furnace to produce pig iron (1.355 × 10⁶ t/year capacity);
• a plant to produce lime (9.1 × 10⁴ t/year capacity);
• a plant for steel making, with two LD converters (135 t capacity each) and three four-strand continuous-casting machines; and
• a complex with a billet mill (7.9 × 10⁵ t/year capacity), a medium-section structural mill (5.6 × 10⁵ t/year capacity), a light-section mill and bar mill (4 × 10⁵ t/year capacity) and wire-rod mill (1.3 × 10⁵ t/year capacity).

A series of protracted negotiations regarding the construction of the plant then started between NSDA and TPE. Negotiations were frequently deadlocked, and a final contract emerged only after a high-level delegation (led by the Chief of Staff, Supreme Headquarters, General Musa Yar'Adua, then the de facto Vice President under the military government of General Obasanjo) undertook a journey to the Soviet Union and met with high-level Kremlin officials. Finally, in July 1979, a contract was signed with TPE for the preparation of working drawings, supply and installation of equipment, and construction of special steel works for the Ajaokuta steel project.

In 1979, NSDA was broken up into various organizations, including ASCL. In 1980, a contract for civil engineering was signed with Western firms: Wimpey was to build a metallurgical training complex and a river port; and project management was awarded to Pan African Consultancy Services Ltd (PACS) and Metallurgical Engineering Consultants (MECON) of India. Construction work began the next year, but within 2 years it came to a virtual standstill, and the partners began to negotiate and attempt to reactivate civil works and set up a more realistic contract.

Other steel projects

A delegation from Nigeria toured various countries to see the types of direct-reduction plants in operation and assess their performance. Following this, the government decided to establish two direct-reduction plants: one at Port Harcourt and the other at Warri, each with a capacity of 5 × 10⁵ t/year. One of the plants was finally shelved, and in October 1977, an agreement was reached between the federal government and a consortium of 10 German and Austrian firms to construct a plant to produce liquid steel at a capacity of 1.0 × 10⁶ t/year. Part of the steel would be finished in the plant's 3 × 10⁵ t/year light-section mill, and the remaining steel would be rolled in major market centres in the country.

In 1979, contracts were signed for three rolling mills, to be built at Katsina, Jos, and Oshogbo. The Katsina plant was constructed by Kobe of Japan, and the Oshogbo and Jos plants were constructed by German companies. Each of the plants was designed to produce bars and wire rods at a capacity of 2.1 × 10⁵ t/year. These were to be rolled from billets produced at Aladja.

The Delta steel plant was commissioned 29 January 1982, and the rolling mills were completed on a turnkey basis.

The 1981–1985 development plans also included the following:

• a flat-products plant (capacity, 1.5–2.0 × 10⁶ t/year) to produce steel to be formed into sheets, plates, coils, thin plates, and welded pipes;
• an alloy and special steels plant (capacity, 8–18 × 10⁴ t/year);
• an iron and steel foundry complex; and
• an aluminum smelter.

The strategies and sequence adopted may be examined with the benefit of hindsight. The initial options can be summarized as the establishment of mini steel mills using scrap and an electric-arc furnace; rolling mills; and an integrated steel works.

The proposal for mini steel mills was questionable because the local supply of good scrap was poor and the infrastructure couldn't support steel making by an electric-arc furnace. However, as such plants usually have a low capacity, the experience gained in establishing and operating them could be expected to offset some of the costs resulting from a lack of economic feasibility.

The rolling mills were scheduled for completion up to 3 years before the primary plants, and when they were completed, they would have to operate at less than 20% capacity because of the lack of steel available to roll. The total capacity for long products was to be 2.36 × 10⁶ t (including private rerollers), although projected demand (Project Review Committee 1981) for long products was 7.5 × 10⁵ t/year and for flat products was 8 × 10⁵ t/year.

The long gestation period and huge escalation in costs for these overly ambitious projects turned the steel dream into a nightmare. Today, the federal government of Nigeria owns two integrated steel plants: Ajaokuta, using the conventional BF–BOF process, and Delta, using the DR–EAF technology, with a combined installed capacity of 2 × 10⁶ t/year. The government also has three inland rolling mills, with a total capacity of 6.3 × 10⁵ t/year; the raw materials for production are to be billets from the uncompleted integrated plants. The rolling mills compete with the private steel mills (Table 1). Even if finances had not dried up and all projects had been completed as planned, the available labour force capable of the requisite managerial and technical-support functions were too few to absorb the technology.

Table 1. Location and installed capacity of privately owned rolling mills in Nigeria.
Firm	Location	Capacity (104 t/year)
Qua Steel Universal Steel Continental Iron and Steel Sels Metal Federated Mills Allied Steel General Steel Mill Nigerian–Spanish Engineering Mayor Engineering Oro Steel Kwara Commercial Metal and Chemical Industry Union Steel Asiatic Manjarin Industries Niger Steel Metcom (Nigeria) Ltd Others (estimated)	Eket Ikeja Ikeja Ikeja Ofa Onitsha Asaba Kano Ikorodu Ilorin Ilorin Ilorin Ikorodu Enugu Owerri	10 8 15 10 14 10 5 19 29 ?? 4 ?? 6 3 3 2

Factors deserving more consideration during the planning stages included the following:

• the stage of technological advancement and all available options in the field;
• the pool of trainable or usable personnel;
• the demands and limitations of the market to be served by the product;
• the effects of the gestation period and the economic climate;
• the scale, complexity, and nature of the processes; and
• the availability of raw materials and infrastructure (Table 2).

The nature, quality, and availability of raw materials used in steel plants affect decisions on technology. In fact, input materials and the choice of technique are more or less inseparable. For example, wheras natural gas is the dominant energy source in the DR–EAF process, it finds only an auxiliary use in the BF–BOF process, which uses coking coal as the major fuel. Nigeria has fairly large deposits of iron ore but a paucity of coking coal. The coal available is riddled with ash and sulfur. The available iron ore can be beneficiated fairly cheaply to meet the requirements of Ajaokuta's blast furnace, but Delta's direct-reduction plant demands a high ferrous content (about 66%) and requires ore with an iron content of 38%. The cost is high. Imported ore may be cheaper than local ore because of the expensive beneficiation and imported coking coal. The fuel for Delta's direct-reduction plant is natural gas, which is so abundant in Nigeria that large portions of the associated reserves (gas that occupies oil wells) are still being flared. The main reserves are practically untouched. Thus, on one hand, the plant that is operable with a cheap fuel may have to rely on imported iron ore; on the other hand, the plant this is operable with local ore is doomed to using imported fuel (coal). The raw materials are about four times the value of the product, and Ajaokuta alone would have to process more than 5.0 × 10⁶ t/year raw materials for its production of 1.3 × 10⁶ t/year.

Nigerian iron ore falls into two categories. The first is hematite–magnetite iron ore, which has been the most explored. A reserve at Itakpe, 66 km from Ajaokuta, is estimated to have 300 × 10⁶ t, with an average iron content of 38%, which can be beneficiated relatively cheaply to about 63% ferrous iron. The iron ore at Ajabanoko, Chokockoko, and Agbaja is similar to Itakpe ore, with a reserve of 150 × 10⁶ t. The second category is an oolitic sedimentary type. A deposit estimated at about 3000 × 10⁶; is located in the Agbaja–Lokoja–Koton–Karifi area. It contains high levels of phosphorus and zinc and has been investigated very little, beyond initial feasibility reports.

The two known coal deposits in Nigeria are the Enugu and Lafia–Obi coal fields. The Lafia–Obi coal has coking properties but is high in ash and sulfur, and the deposit has structural problems. The coal in the Enugu deposit, on the other hand, is reasonably free of impurities but is noncoking. Thus, it was decided that, initially, the coking coal would be imported. However, technologies exist to make at least 20% of a fuel blend from Enugu coal. The savings would be significant. Another deposit, at Okaba, has coal with high ash content and low tar yield, and this coal disintegrates on carbonization, although its use in the production of steel has not been ruled out.

Limestone is supplied to Delta Steel by Cross River Limestone at Mfamosing. Delta has equity in the deposit. A deposit at Jakura has been proposed for Ajaokuta because of its proximity to the plant, but Jakura limestone would have to be formed into briquettes. The government has not taken a definitive stand on the use of coal and most of the other materials and is still deciding whether to award contracts for the supply of briquetting machines.

Deposits of dolomite (4 × 10⁶ t) in Burum, near the federal capital, are also not far from the plant site. Both deposits need to be fully explored, and the issues of ownership and mining rights need to be resolved. Refractory deposits are found in Oshiele–Onibode (very far from Ajaokuta) and in Ozubule, Imo.

Both bauxite and ferric alloys are to be imported, although low-grade manganese (2 × 10⁶ t) has been reported in Tundun Kudi, northern Nigeria, and bauxite has been reported in Oju, Benue.

The Ajaokuta dream

To prepare coke, the plant has 49 ovens for each of two oven batteries (5.5 m high, with a useful volume of 30.3 m³). The capacity of the batteries is 9.0 × 10⁵ t/year. The sintering plant has two machines that produce 100% self-fluxed sinter. The plant is designed to use iron ore from the Itakpe mines.

Iron making is done by a single blast furnace, with 2.0 × 10³ m³ capacity. It has been adapted for natural-gas injection, to reduce coke consumption. The pig iron produced (1.5 × 10⁵ t/year) is used in foundries; the remaining molten iron is sent to the steel shop; and the slag produced as byproduct (5 × 10⁵ t/year) is used in cement making.

The steel-making shop consists of two LD converters with 130 t capacity each, three two-strand continuous machines, and the lime shop. There are four rolling mills: a 320 mm light-section mill; a 700 mm medium-section structural mill, a 900/630 billet mill, and a 150 mm wire-rod mill.

The greenfield nature of the plant, the unreliability of public utilities, and the dearth of small-scale suppliers in the immediate vicinity plant meant that the development of the steel works had to include special provisions for spare-parts manufacturing facilities and power plants, etc. Among the facilities needed for Ajaokuta were a comprehensive repair complex, with its own foundry, forge shop, and heat-treatment and hard-surfacing shop; a thermal power plant and turbo blower station; an oxygen plant; refractory shops and a lime plant; and laboratories and transportation facilities.

Also needed were external infrastructure, such as railways, roads, ports, and electric-power systems. For example, there was a proposal to construct a line from Onne port through Port Harcourt to Ajaokuta, although it was shelved for financial reasons. A line between Itakpe and Ajaokuta was to carry iron ore, and the line running from Lafia–Obi and Makurdi was to carry local coal. Onne port was to be built to handle imported raw materials and finished products for Ajaokuta, but with the cancellation of the rail program, the use of Onne port was questionable. Electric-power facilities included a 330 kV double-circuit transmission line, from Benin to Ajaokuta, and the captive electric-power system at the plant site in Ajaokuta.

Planning the plant at Ajaokuta

Table 3. Estimated capital and operating costs for steel plants at three possible sites in Nigeria (1984).
Ore supply	Capital costs (106 NGN)				Operating costs (106 NGN)
	A	B	C	Total
Imported and local Warri Onitsha	504 341	182 142	142 126	828 609	187 143
Local Ajaokuta Onitsha	359 331	231 142	158 126	748 599	201 163
Source: NSDA (1974). Notes: A, transport, water, electricity, and gas; B, substructure work, site leveling, construction of plant; and C, construction of township for the main contractor and the Nigerian personnel. In 1995, 78.5 Nigerian naira (NGN) = 1 United States dollar (USD).

Two suitable zones were defined: the coastal zone, stretching for about 100 km between Forcados and Port Harcourt; and the central zone of the country, along the banks of River Niger, south of Lokoja, for 50 km to Onitsha.

Eleven possible locations in these areas were distilled down to three: Warri, Onitsha, and Ajaokuta. The report recommended Onitsha, on the premise that the use of local ores would save precious foreign exchange.

Political factors outweighed the economic considerations, although the reasons for this are a matter of conjecture. The preliminary report was submitted for consideration in 1974. This was 3 years after a civil war, when the eastern zone of the country (with Onitsha its trading capital) attempted to break away and set up a separate nation. With the memories of the war and its scars still fresh, the authorities might have hesitated to choose to site a project of such huge cost at Onitsha: they would be unable to guarantee the security of government property or the safety of workers from other regions of the country. In other words, the decision was economically suboptimal but, perhaps, politically justifiable. We suggest that it is no use pretending that noneconomic factors are irrelevant. In fact, they tend to be decisive in developing countries, so policy analysts must seek to accommodate them.

In contrast, the choice of technology seemed economically sound. The main groups involved in this decision were the main contractor (TPE of the Soviet Union), the Nigerian engineers (from NSDA), and the consultants to NSDA (Sofresid of France). The preliminary report by the Soviets considered two options for the iron-making plants: two blast furnaces, with an effective volume of 1003 m³ each, and a single blast furnace with an effective volume of 2000 m³.

Sofresid and NSDA engineers agreed that either option would meet the requirements for hot-metal production. The preliminary report pointed out that labour efficiency with a single furnace is higher by about 30%, and the prime cost per tonne of hot metal was about the same for the two options.

Nevertheless, the report favoured the two blast furnaces, to avoid complete shutdowns when repairs were needed. Sofresid observed that the equipment proposed in the report was obsolete, by Western standards, and recommended areas for improvement.

The final decision was in favour of a single 2000 m³ blast furnace, capable of producing 1.355 × 10⁶ t/year of hot metal and incorporating high top pressures, as well as auxiliary fuel injection. Some of the arguments that favoured the single, large furnace cited an expanding market for steel products, constraints on the amount of available land, and a high rate of capacity utilization. A major drawback of the equipment chosen was its need for 100% sinter as raw material. Small furnaces are more tolerant of poor-quality raw materials (i.e., more local coal could have been incorporated in the coal blend) and are easier for inexperienced operators to manage. Mistakes are easier to correct, and reliance on instrumentation and controls is less critical. In the Nigerian environment, pure scale effects constitute a less valid reason than furnace availability and operability.

NSDA engineers made their decision without the benefit of hindsight or the technical skills that accumulate from operating such plants. The engineers believed the national economy would remain buoyant enough to finance an expansion in capacity immediately after the completion of the first phase, and they also felt that the blast furnace of the first phase would be a training ground for the subsequent phases.

However, the economy collapsed and the expected growth in demand for steel products evaporated. Clearly, decision-makers must build in some allowance for uncertainty and change. ASCL probably cannot finance even the coke it requires (150 million USD/year), under present economic conditions.

Uncertainty derives not only from the economy but also from the technological environment, particularly the weak capability to maintain equipment on a sustained basis and the lack of local supplies of spare parts and machines. Learning was taken for granted and was not given serious attention in decision-making. The project's evaluators made no attempt to analyze the potential opportunities for gains in technical know-how offered by the two options, and few if any long-term social objectives were considered in the decisions.

The steel-making shop proposed in the preliminary report consisted of two oxygen converters, each with a 130 t capacity. This proposal was accepted because the concept of oxygen steel making was well established. Less clear cut was the decision about the process for casting. Traditionally, molten steel is poured into ingot moulds, allowed to cool, and then reheated and shaped. A more recent technique — continuous casting, or concasting — casts the molten steel directly into blooms (typically, 260 mm × 260 mm or 335 mm × 6 m), billets (100–150 mm² × 12 m), or slabs. Concasting increases the yield of cast and has great advantages in productivity, material handling, and fuel costs.

The preliminary report recommended the use of concasting, and Sofresid agreed. However, the consultants did not agree on the product. The Soviets recommended blooms (with later reduction to billets in a mill) and slabs. Sofresid suggested casting billets directly because it offered a reduction in capital and operating costs. It is the not clear why the Soviets recommended a bloom caster, but it is suspected that the technology of billet casting was still new to Soviet steel-plant designers; hence, there would not have been a manufacturer from the Soviet Union to supply the plant.

The Soviet proposal had following advantages:

• It is easier to cast blooms than billets.
• It is possible to produce variable sizes of billets in a mill, thereby increasing flexibility for sales (a billet mill is a simple rolling mill with few stands and, thus, is not likely to create a production bottleneck).
• Natural gas for the operation of reheating furnaces is locally available, so costs for converting blooms to billets would be a local expenditure (except for imported rolls and accessories).

The evidence indicates that the Nigerian engineers carefully considered the technical and economic factors and decided in favour of bloom casting. Casting the billets directly would have been preferable, as this would have resulted in some standardization in the rolling mills and would have eliminated the need for a billet mill. The decision of the NSDA engineers could have been influenced by their training in Indian steel plants, where continuous casting facilities were not yet available.

For the rolling mills, the preliminary report proposed a product mix consisting of equal amounts of flat and long products, a total of 1.3 × 10⁶ t/year. This was based on the current demand. However, at the time, the national economy was booming, and the construction industry was enjoying higher growth than the manufacturing sector. Large office blocks, industrial complexes, and long bridges were being built, and the government had plans for heavy industrial investment in liquefied natural gas, petrochemical plants, and fertilizer plants. All of these would require large amounts of steel sections, mainly long products. This led to the decision that the first stage of the plant would be devoted to long products, and the second stage — an expansion to 2.6 × 10⁶ t, which was expected immediately — would be devoted to slabs that could be rolled into flat sheets.

The preliminary report recommended

• an 800 mm billet mill,
• a 450 mm medium-section structural mill,
• a 250 mm light-section and wire-rod mill,
• a 1700 mm hot strip mill,
• a cold rolling mill,
• a corrugated-sheet mill,
• a hot-rolled coil-cutting unit,
• an electrolytic tinning unit,
• a pipe spiral-welding plant, and
• a pipe electric-resistance-welding plant

Hindsight shows that the change of the original concept of the plant was a grievous error. The overall interest of the country would have been served best by a mixture of flat and long products. Today, Nigeria is flooded with long products, and every industrial concern voices the demand for flat products. The refrain has led the government to adopt what it calls "accelerated phase of expansion — flat products stream." What it proposes is to add a slab caster, a hot mill, and a cold rolling complex, with associated auxiliary facilities, to produce hot-rolled and cold-rolled products. The new setup would take about half the molten steel from the units of the first phase.

Optimism about future economic growth was probably also to blame for the inclusion of a medium-section structural mill, which was done at the suggestion of NSDA. The contractor noted that demand for medium sections was low and recommended, instead, a section forming unit. A forming unit offers higher efficiency and lower capital cost than a medium-section mill. It is, however, slower and not amenable to high-tonnage production.

The NSDA engineers decided in favour of a high-capacity, medium-section structural mill to anticipate growth in demand for its products from heavy industry, including the plant itself. At advanced stages of project execution, the final design of the rolling mill was further modified, after the detailed project report, to incorporate production of rails.

The maintenance philosophy put forward in the preliminary report, at the suggestion of Nigerian engineers, was a central system of repair facilities. Thus, the report contained elaborate provisions for machining new spare parts and consumable equipment in departmental workshops. Provision was also made for large repairs — an iron and steel foundry, forge and fabrication shop, machine shop, and a shop for heat treatment and hard surfacing. Sofresid questioned the rationale for such enormous facilities, saying that (1) these provisions were excessive for a steel plant, by usual standards; and (2) the plan featured duplication and overequipping of the departmental shops.

The shop featured excessively large cranes. The repair shops for power equipment would be able to repair of 13 000 electrical items annually. The estimated requirements for spares at the facilities were astronomical by any standards (Table 4). Also featured were a rubberizing workshop, an equipment storage and charging station, a block of workshops dedicated to the steel-making plant, a workshop for the rolling mills, and a workshop for the coke-oven and byproduct plant.

Table 4. Estimated requirements for spare parts in repair shops.a
	Requirement (10³ t/year)	In-plant production (10³ t/year)	To be bought (10³ t/year)
Iron castings	4.33	4.23	0.10
Steel castings	4.58	4.48	0.10
Nonferrous castings	0.19	0.19	—
Forgings	4.48	2.28	2.20
Long products	2.10	—	2.10
Steel structures	2.50	2.20	0.30
Cast-iron rolls	0.86	—	0.86
Steel rolls	1.60	—	1.60
Copper plates	0.17	—	0.17
Machining of new parts and consumable equipment	8.36	8.29	0.07
Remachining for reclaiming and rehabilitation	1.43	1.43	—
aPlant design capacity with a two-shift operation.

Repair shops were to supply 67% of the spare parts, although some were to be specialized products that could be economically produced outside dedicated plants. That the work, according to the report, represented a two-shift operation suggested that a 50% increase would be possible with three shifts.

In engineering circles, Soviet equipment is regarded as generous in its allowance for capacity. It is large and electronically unsophisticated. The Soviets make allowances for weak industrial cultures by building in extra capacity. For example, a Soviet contractor will supply a 2500 t/day furnace to a developing country, labeling the equipment as a 2000 t/day furnace. This is safe for the contractor, as it is then easy to demonstrate the guaranteed performance, and plant operators are able to meet design capacity easily. This practice has endeared Soviet technology to several developing countries, where most steel plants have remained wholly government owned.

Overdesign is not necessarily negative. It may serve as the basis for growth and minor innovations. For example, Maxwell (1982) observed the preponderance of overdesign in some Latin American steel plants and noted that initial imbalances removed the need for investment in new capacities during expansion.

The elaborate provision for spare-parts manufacturing was aimed at developing a local capability for repair and replacement of worn parts and manufacture of whole machines, where necessary. Developing the capability to effectively use these plants would lead to the independence of the plants from outside suppliers and would be a base for manufacturing parts (in case the contractors' equipment needed to be amended to adapt to specific local conditions or to increase the productive capacity).

What, in an economist's view, is overcapacity could, in the long run, provide the financial buffer that enables the steel plant to break even. The facilities already have received a deluge of manufacturing orders from outside firms.

Dahlman et al. (1978) documented how the repair shop at the Usiminas steel plant in Brazil grew into a successful manufacturing concern through systematic investment in human and material resources. Usimec (the machine-making subsidiary) designs and builds all the capital goods Usiminas needs and also serves outside consumers. What Usimec is doing for Usiminas, the Growth Shop is doing for Tata Iron and Steel Co. in India.

Such commercial endeavours are possible because local suppliers cannot support industry, which is underlined by the experience at the Delta steel plant. Six years after commissioning, Delta (which is nearer than Ajaokuta to an established industrial environment) still cannot obtain rudimentary spare parts and consumables locally. Great foresight was not displayed in the design of the repair facilities. The Soviets favoured and adopted the repair-shop approach, probably because of their experience in supplying steel plants to developing countries, whereas Sofresid, the French consultants, found the facilities superfluous. In France, it may not be cost effective for a firm to be unburdened with such provisions, which in themselves may pose organizational problems.

Thermal power plant

The preliminary report recommended three 55 MW generators, fueled by oil or natural gas. Sofresid recommended an alternative: three 110 MW generators that would generate even more electricity than the national grid system and would make the steel plant independent of the grid. The NSDA engineers opted for two 55 MW generators. The peak power demand is 220 MW. The in-plant power generation was based on the available byproduct fuel and the minimum power needed to carry the critical equipment in the plant.

The two independent sources of power are a technical requirement, although some agencies contend that both sources could come from the National Electric Power Authority (NEPA). This argument is flawed by the fact that NEPA has never been reliable.

However, running a full-size 110 MW power plant places a heavy load on the technical and organizational management of the steel plant. The organization has to cope with power generation (a large, complex plant) a spare-parts-manufacturing complex, and large waterworks, as well as the integrated steel complex. Such diverse activities require a very able technical administration. The integrated plant includes facilities with different technical characteristics. For example, the coke-oven byproduct plant is like a petrochemical complex and miniature fertilizer plant; the links among iron making, steel making, and continuous casting are sensitive to, and intolerant of, poor planning and sloppy management. Because of the managerial competence required, some form of goal-oriented decentralization seems appropriate, although that may also be a source of problems. Where government allocation of finances to projects can be affected by pressure groups and rapidly changing priorities (because of political instability), one section of a project may proceed smoothly while another essential component under a separate management organization experiences delay.

A good example is the development of the Itakpe mines: the contracts for the ore beneficiation plant and the Ajaokuta–Itakpe rail line were not awarded until 1987, 4 years after the blast furnaces at Ajaokuta should have been commissioned, and the contracts required a minimum execution time of 22 months. How developing countries, with their limited resources, can avoid such incongruencies is unclear. The choice seems to be either one large organization, with complex and diverse technical arms under a single umbrella, or a number of independent smaller organizations, trading services with each other but having a central coordinating agency.

Major actors and their roles in the Ajaokuta project

1. The development contractor was TPE, the Soviet company that supplied working drawings and structures and installed the process units.
2. The owner, or client, was ASCL, which acted as proxy for the federal government of Nigeria.
3. The contractors for civil engineering work were primarily Fougerolle Nigeria Ltd – Fougerolle SA; Bilfinger and Berger Bauaktiengellschaft – Julius Berger Nigeria Ltd; and Dumez Nigeria Ltd – Dumez Africa. George Wimpey was awarded a contract for the training school complex; Boskalis was contracted for the river port; and two of the civil works contractors (Julius Berger Nigeria Ltd and Fougerolle Nigeria Ltd) were given contracts for infrastructure.

The major obligations of TPE, the development contractor in the Ajaokuta project, were the following:

• preparation of working drawings for all the construction, erection, and special civil works;
• delivery of the equipment, steel structures, refractories, rolled stock, pipes, and other articles and materials required to carry out the erection and special civil works at the plant units and the erection base;
• erection and special civil works at the plant units and the erection base, as well as training of Nigerian workers to master the erection specialities;
• supervision of start-up and commissioning of plant units;
• provision of technical assistance during the operation of the plant, as spelled out in a supplementary contract; and
• training of Nigerian personnel to operate the plant.

ASCL's main obligations for the execution of works were the following:

• preparing the units and the site for erection, including readying storage areas for equipment, steel structures, pipe, and refractories, etc;
• undertaking civil works for the construction of the erection base;
• ensuring availability of river and sea ports and river-port facilities;
• building upland and lowland canals;
• providing offset water supply, sewage treatment, rail and roads, and natural gas pipelines;
• ensuring availability of national and international communication facilities;
• establishing the township for the main contractor and Nigerian personnel;
• investigating further sources of raw materials (coal, limestone, dolomite, refractory clays, and iron-ore concentrate) and submitting the results to the contractor at a stipulated time (to be determined later);
• coordinating, managing, and supervising construction of all facilities in cooperation with the Soviet experts who were supervising the civil works to guarantee quality of work; and
• providing local personnel for erection work.

Details of the civil contract were finalized in September 1980. The broad schedules for construction of civil works and erection of steel structures and equipment were agreed on by ASCL and TPE during a series of meetings in Moscow, lasting from December 1980 to February 1981.

Delays

Table 5. Division of lots for civil works contracts in the Ajaokuta steel plant project.
Lot	Contractor	Civil works
I	Fougerolle Nigeria Ltd – Fougerolle SA: joint venture	Raw material plants, coke ovens, blast furnace, thermal power plant
II	Bilfinger and Berger Bauktiengellschaft – Julius Berger Nigeria Ltd	Steel-making shop, rolling mills
III	Dumez Nigeria Ltd – Dumez Africa: joint venture	Auxillary shops, lime and refractory plants

The major reason for the civil contractors' deceleration and eventual demobilization was that funds earmarked for the project were exhausted. When the contract was negotiated, it was anticipated that the economic landscape would be unstable, and a cost escalation factor was included in the agreement. However, wages and salaries were based on prices in 1981, and no allowance for increases was made. Sociopolitical and economic forces were already at work to make a mess of the calculations. A new civilian administration was voted into power at the end of 1979. As part of the new government's economic program, it revoked the system of "zooming," whereby skilled, semiskilled, and unskilled labour were paid at different rates. As a result, wages increased 196–384% in some categories of labour.

The world economic recession, with its concomitant effect on the Nigerian economy, led the government to introduce the Economic Stabilization Act, which caused sharp escalations in the cost of raw materials (both local and imported), transportation, and services. The civil contractors used enormous amounts of cement and steel structures purchased locally and abroad and hauled large quantities of these materials over great distances, which seriously affected their costs.

Also, the March 1980 call for tenders for civil contracts left out main units and off-site facilities needed for the general operation of the plant; for example, the permanent water-supply intake, water-treatment plant, general-purpose drinking-water supply, sewage-treatment plant, sludge disposal system, and plant boundary wall were not included because the design date and specifications were inadequate. Similarly, construction of low-cost housing for the civil contractors and of site facilities subsequently added to the cost of civil works. Also not included in the 1980 call for tenders were the river port and the metallurgical-training complex. These eventually ballooned costs.

During construction, changes are usually made: equipment is added to strengthen the original designs; excavations and earth fillings deviate from original estimates; and design errors compel modifications and, therefore, substantial cost revisions. Some examples from Ajaokuta are the following:

• the transfer of work from Dumez to Bilfinger and Berger;
• the decision of the federal government to modify the medium-section structural mill to make it produce rails; and
• modifications to the 17 m × 17 m × 1.5 m foundation of the blast furnace (the foundation required continuous pouring of concrete for more than 36 h).

By March 1983, the government was directing ASCL to negotiate with the contractors to arrive at a realistic and equitable formula for price fluctuation. Protracted negotiations followed, and after more than 2 years, a formula of 1.5 was fixed. For further escalation in costs of material and labour, a system of basic rates was fashioned, with the base rate being the one arrived at in April 1984. The standstill and the negotiations saved the project about 125 million NGN, but the costs in time overruns and lack of production cannot be easily calculated (in 1995, 79.5 Nigerian naira [NGN] = 1 United States dollar [USD]).

The civil works were expected to take 54 months from November 1980. More than 70 months elapsed, and much still remained to be done. Meanwhile, the new agreement, estimated at 838 million NGN, catapulted to 1484 million NGN.

After the new agreements and new commissioning dates were established for the plant units, other external threats — the shackles created by bureaucracy and the uncertainties in the economic climate — emerged to make the new timetable unrealistic. The new commissioning dates were predicated on the assumption that import licences for the contractors (for the 1985/86 financial year) would be available by August 1985. In fact, the licences were not in place for more than a year. The civil contractors, who were expected to resume work once the deadlock was broken, had not done so by late 1986, partly because the new mode of payment for the civil works was being processed through the bureaucracy and partly because the import licences had not materialized.

The contractors resumed work towards the close of 1986, but before the end of first quarter of 1987 new problems relating to the mode of payment had brought the work to a standstill once again. Paradoxically, the abrupt halt to progress on the civil works made little difference overall. Infrastructural facilities, external to the plant, are meant to supply, process, and convey raw materials to the plant. Even if the steel plant itself had been constructed on schedule, delays in the installation of these facilities would have rendered it inoperable. The most important delays occurred in the following:

• the Itakpe iron-ore mine, which produces the concentrates to feed the Ajaokuta plant, 56 km away;
• the railway link between Itakpe and Ajaokuta to transport iron ore, the contract for which was awarded only in 1987;
• the development of mines or quarries for limestone, dolomite, and refractory clay and of transport facilities to convey the raw materials to Ajaokuta; and
• the development of a bulk transportation system to carry imported raw materials, like coal, from Onne port to Ajaokuta.

The direct and indirect costs of delay were many:

• Capital was tied up in half-completed civil works, partially installed equipment and plant units, uninstalled structures and equipment, and idle tools and equipment for erection.
• Workers were demobilized and remobilized.
• Equipment needed refurbishing. With time, conservants like grease and oils that protect equipment from rust lose their potency; of about 1.8 × 10⁵ t of equipment, half is expected to need reconservation.
• Installed but idle plant units needed protection from a hostile environment.
• Completed but idle units depreciated in value.
• Supplies for plants and equipment, erection, and commissioning were deferred (if ASCL could not supply the requisite materials, it would be required to pay highly for the time the contractor's personnel were idle on the job).
• Interest accrued on unproductive loans (100% of the civil-works contracts were externally financed, and the bulk of the money was to be paid back in foreign exchange).
• Prices escalated.
• Workers were idle, and morale was low. There was a large turnover of skilled and unskilled workers waiting for the plant to come on stream (in some instances they waited for more than a decade after their formal training in steel-plant operations). The firm lost several experienced engineers and administrators for political reasons (mass purging); others left for better pay in the private sector or out of sheer frustration).
• Output and revenue were unrealized; the plant failed to contribute to the economy of the country.
• Technical assistance was required.

The major factors causing delays were the nature and complexity of the project. The decisions made and their subsequent implementation demanded unusual technomanagerial capabilities. Because of the paucity of these kinds of capabilities in Nigeria, it is not surprising that the project faltered.

Practically all the capital goods — structures, columns vessels, and reactors — had been made abroad, the products of detailed knowledge acquired from many years of design and fabrication. Not being a party to the design and fabrication is an operator's first major handicap. Understanding the technical basis for a design limits guesswork and enhances the operator's ability to make judgments about use and maintenance. Clearly, the project demanded massive investment in training and development.

Personnel training and development

Training was mainly ad hoc, lasting from 3 to 9 months, and NSDA simply took what was offered. The contents of the courses varied and were usually predesigned. Participants supplemented these learning opportunities by regularly reading journal articles on iron and steel (NSDA maintained an up-to-date library on steel manufacturing). Some of the ad hoc arrangements continued until 1974, when several senior staff of NSDA were sent to MECON (India) to study general layout and transport, civil engineering design, structural and project engineering, and so on.

Training in India

The training program was designed entirely by SAIL, with each trainee spending 9 months on specialized work and 9 months participating in shift work. The contract between SAIL and NSDA was valid for 4 years, starting August 1974, and the scheme involved mainly executive and engineering staff. The experience of early trainees resulted in two major amendments: the period was reduced to a total of 12 months, and training in the use of some critical equipment, like cranes, was included. The intention was to equip executives and engineers with skills that would enable them to keep vital operations going and to judge the performance of workers.

The training of steel-plant personnel was conducted at the Bhilai, Bokaro, and Raurkela steel plants, Coal India, and the National Mineral Development Corporation, among others. Some design training took place at MECON's design office in Ranchi.

The trainee's program was divided roughly into three phases: (1) general training, (2) specialist training, and (3) management training.

In the first phase, the trainee was taken through all the activities of the steel plant, from raw material receiving, handling, and preparation to the finishing end, in this case, the rolling mills. In the second phase, the trainee then spent some time interacting with the engineers, technicians, and shop hands in specific operations or maintenance. Next, the training focused on, for example, coke ovens, byproducts, or steel making. Shift work was mandatory. The third phase, the general management course, lasted 2 weeks. The course was designed to familiarize trainees with basic accounting; project implementation; production, materiel, and personnel management; marketing; and safety.

Another contract with MECON proposed training another 636 engineers, staff operators, and nonexecutives, but this never got off the ground.

Training in the Soviet Union

The program started in 1981. The first Nigerians trained in the Soviet Union were mostly trained for the priority rolling mills (light-section and wire-rod mills), as well as for the operation of the thermal power plant, the gas facility, and the electric power facility.

A new schedule calls for training fewer people abroad; only those needed for the most critical categories will be sent. A "cascade" approach is now planned: those already trained will be expected to impart their knowledge to others in Nigeria, mainly at the technician–operatives level.

According to the original contract, training was to be carried out "in conformity with the training programmes prepared by the establishments where training will take place." This seems to reflect the climate of the time, a time when NSDA did not have the personnel, the confidence, or the capability to plan such a venture. The direct experiences of the NSDA students resulted in revisions to the contract, such as shortening the length of training and focusing on training in critical equipment.

There was a tendency to leave the content of the courses up to the Soviets. For example, the contractor was to assign trainees "to the various establishments taking into account the relevance of these establishments to the training programmes, the similarity of the available equipment to that to be supplied to purchase."

The contract stipulated that "industrial and technical training of trainees shall be conducted in the Russian language. However, to ensure more efficient training, PURCHASER'S trainees shall, prior to coming to the USSR [emphasis ours] for industrial and technical training, study the required minimum of the Russian language in PURCHASER'S country."

The obligation to ensure that trainees have a knowledge of the Russian language has been treated with levity. When trainees get to the Soviet Union, a language program, not exceeding 3 months, is normally arranged, along with the specialized training. Not surprisingly, most trainees never really understand the language. This imposes serious constraints on learning. As most texts are printed in the Russian language, trainees who wish to explore technical books and journals find it is a barrier as much at the library as on the shop floor. Invariably, they misinterpret technical information, as do the interpreters, who are language specialists, rather than technical specialists. The communication gap sometimes leads to long, unproductive periods of clarification and repetition — time that could be used more productively.

Training in Nigeria

To meet this requirement, the metallurgical-training complex was to be built to train about 200 at a time, in 25 different trades and skills. Graduates would number about 900 annually. Fully equipped classrooms and workshops for practical training, as well as hostel facilities, were to be provided. The current complement of staff at Ajaokuta comprises 666 engineers and technicians and 1163 other workers (Table 6).

Table 6. Personnel working in various departments of the Ajaokuta steel plant (1987).
Department	Engineers and technicians	Other workers
Coke oven Steel-making plant Iron-making plant (including blast furnace) Maintenance (including lime plant, refractories) Power and utilities Central repair-shop complex Research and quality control Production, planning, and control Light-section mill Wire-rod mill Billet mill	18 21 18 95 173 108 47 17 50 49 70	4 79 29 190 205 76 14 3 200 208 155

However, the thinking that impelled an investment in a massive training institution made no provisions for the future, when the knowledge gap will have been narrowed. A survey carried out in 1973 (and updated in 1976) indicated that certain skills simply could not be met by educational programs in the country. At that time, metallurgy, metallurgical engineering, mining, refractory technology, and instrumentation and control were not taught in any Nigerian institution. These findings prompted the initiation of a postgraduate diploma course and a master's degree program at the University of Lagos. The aim was to give chemical engineers, chemists, and mechanical engineers the basic concepts of metallurgy, with emphasis on physical and extraction metallurgy.

Two of the rolling mills are now in operation and are overstaffed by engineers and technicians and critically short of other kinds of workers (UNIDO 1984). As a result, personnel trained to run the coke oven and steel shop have been converted overnight to mill operators. At the coke-oven plant and steel-making shop, on the other hand, there is a shortfall of engineers who have received their local and Soviet training. Both the coke-oven plant and the steel-making shop have yet to go through the commissioning stage, a stage when staff participate in equipment start-up and dry test runs. The familiarity staff would develop with plant equipment prior to commencement of commercial operations could be crucial to the efficient operation and development of a no-nonsense maintenance crew. At present, engineers are underemployed, and many have already left.

The strategy of cross-posting of trained personnel from plants under construction to completed plants arose mainly for financial reasons (and, partly, for reasons of organizational politics). The main point is that when the financial resources committed to large projects of this nature start to dwindle, training is the first program to suffer. This has a crippling effect, leading to the prolonged infancy of the firm.

Some other shops in the plant are fully constructed, yet they too are constrained by shortages of trained personnel. The most striking example is the central repair-shop complex. We can judge from the experience of other plants, such as Usiminas (Brazil), that this complex should be fully exploited to serve potential customers. The shop requires 115 engineers, but no one has been formally trained. This shop is incontrovertibly the most complex and best equipped in sub-Saharan Africa. It has a projected staff of 1677, of which 702 are expected to be highly skilled workers (Table 7). To date, few of these workers have been recruited, although more than 200 machine tools have been fully installed and the shop has been commissioned.

Table 7. Full complement of personnel required for the central repair-shop complex of the Ajaokuta steel plant.
Shop	Engineering staff	Skilled workers	Semiskilled workers	Unskilled workers	Service staff
Foundry Forge and fabrication Mechanical repair Building repair Power equipment	17 12 40 20 26	78 53 254 67 250	48 34 145 82 152	57 38 37 93 106	17 10 15 9 17

The statistics are one measure of the deficiencies. Close observation of the system, as well as direct interviews with various personnel, enabled us to deliver an analysis that goes beyond the statistics. We have several observations.

The recruitment, training, and development of personnel have been inadequate. The numbers of personnel lag seriously the required numbers for both commissioned units and those under construction; it would seem that the plant will face a personnel crisis.

Key units are being commissioned without the participation of carefully prepared, well-trained teams. For large technological complexes, such as the Ajaokuta steel plant, semipermanent working teams should be set up to ensure continuity in the technology acquisition. Included on these teams should be people from the enabling system (government in this case) the ASCL board of directors, and the various departments of the firm).

Political instability — with frequent changes in the leadership of government ministries and in the portfolios of directors, general managers, and senior management staff — makes continuity almost impossible. For example, since 1979, when ASCL was set up, it has had three general managers, not counting the project manager, who retired at the start of construction. For half of this period, ASCL was under direct ministerial control, as there was no board of directors while a new government took the time identifying and appointing its own people. The other half of the time, the management of the firm was shared by two boards of directors. The major disadvantages of direct ministerial control are the bureaucratic delays and the lack of the flexibility needed to efficiently run a corporate manufacturing enterprise.

No one stays in any position or manages any portfolio long enough to sufficiently appreciate the importance of the roles of people in technology acquisition. Under such conditions, it is impossible for people to accumulate the required technomanagerial skills on the job (learning from past errors). Yet, this is the basis for an enduring technomanagerial buildup. This situation is by no means peculiar to Nigeria. Perhaps the way to seek continuity is to assume a system that insulates large technological projects from the weak political culture of developing countries.

One of the prerequisites for the smooth transfer of knowledge is a shared language, yet no serious effort has been made to ensure that language is not a barrier. Interpreters can never substitute for direct communication.

With the advantage of hindsight, we can say it is doubtful the project would have achieved its objective, given Nigeria's recent entry into the steel-producing league, its shortage of experience, and its low level of technological capability. It is also doubtful that a team of the type needed to cope with a plant of the size and complexity of Ajaokuta could have been raised under the circumstances. The project never had the industrial-behaviour pattern or the level of organizational discipline it needed, nor did it have the time needed to achieve these. It may be that there were misconceptions about the technical and managerial know-how available in the country.

The major factors causing time and cost overruns also contributed to the unhealthy state of the training program. A delay in implementation tends to justify (on the surface) shifts in the training program — Why train people when there is such uncertainty about completion date? The learning process, however, is suboptimal and cannot be otherwise when staff's education in the technology does not include textbooks. However, despite an unpreventable atrophy of their knowledge and skills, the staff whose training took place a long time ago form the bulk of management and, by all accounts, have a better appreciation of the issues than those who are completely uninitiated into plant practices.

The requirements for certain basic qualifications, especially those of skilled and unskilled workers, have not always informed the selection process. For parochial and political reasons, personnel have been installed whose abilities are inadequate. Conscious steps should have been taken to ensure that a project of this nature attracted and trained the best and the brightest graduates of the universities and polytechnical institutes. This could have been achieved by open and preemptive recruitment. For example, although the company experiences critical shortages of certain personnel (electrical and electronics engineers and electrical technicians), college graduates are roaming the streets elsewhere in the country, looking for employment.

The chain of learning is broken when newly trained personnel leave for higher paying jobs; yet, the company pays the lowest salaries in the manufacturing industry, and this is a training ground. The accumulation of the required technological skills is clearly impossible under these conditions.

Nontechnical staff in operations and maintenance receive some training, but those in stores management, inventory control, finance, and the training unit receive little or none, although they need a technical understanding of their disciplines.

A lack of confidence has emerged in the firm: this has resulted from overreliance on consultants for tasks that Nigerians have proven themselves capable of doing. For example, the Soviets have delayed handing over the bulk of spare parts and consumables to ASCL officers, with the result that the plant sometimes must wait for parts that were delivered years ago but are locked up in the stores to be documented.

There are no computer facilities to handle the 200 000 drawings and 20 000 t of spare parts, and the study of information storage and retrieval problems has hardly begun. The company realizes the need for computer facilities but has been denied funds or has been caught in bureaucratic delays while continuing to strain under a problem that will require years of expert labour to solve. In the company archives (which stores drawings and various operation and maintenance manuals), there are only a few people with a modicum of training on documentation and registration. The company library — a centre for intellectual activity when the detailed project report was submitted — is now a storage room for old books.

The gradual decline in the calibre of managers responsible for the firm's training program may reflect the true place of training in the management priorities of the firm. At one time, the training managers were highly trained engineers, who were deeply committed to the project and appreciated the role of training in the acquisition of technological capability. Training no longer headed by an engineer: it has been relegated to a small unit that has little role in the scheme of things.

Other crucial segments in the acquisition of knowledge, such as periodic programs for engineers and managers, do not receive adequate attention. Intermediate-level personnel rise to senior management levels without being trained. There is no clear vision of how training and development should be related to appointments and promotions. The unit cannot be so isolated from the system. The role of training in skills acquisition has to be kept in view.

The main functions and roles of a training team are the recruitment and selection of candidates; the intensive and extensive study of the basic-training program contents; the comprehensive study and maintenance of up-to-date information on the human resources of the firm; control of, and accessibility to, facilities required for training and development; monitoring of the training program (basic, in-plant, and post-training or operational-phase performance); and subsequent development and upgrading of personnel as they change roles or get promoted.

The training team currently does not have the ability to manage these tasks, which are essential to systematic accumulation of know-how.

Conclusions

Preinvestment was characterized by the sharing of functions among major contractors (suppliers of equipment), the client, and other participants. The major contractor was commissioned to provide a preliminary report; this was followed by detailed discussions with Nigerian engineers, before the final reports emerged. Events demonstrated clearly that project formulation and execution are complex and multifaceted, demanding an in-depth knowledge of technology and a team of skilful people.

The many issues that called for attention showed how little prepared we were to undertake the project. The sheer magnitude of the problems relating to raw materials and plant infrastructure belied the initial optimism and exposed the fragility of the inchoate organization that was expected to produce steel for Nigeria.

The elements in the project's formulation that shaped the way the project moved were finance (supply and services) and the scheme for implementation.

The civil-works contractors (who were also the suppliers of external credit) dictated the pace of the project. ASCL was expected to survey and explore for all the raw materials, provide port facilities, construct rail and road networks, and so on. A critical assessment of the relative management capabilities, experience, and human resources of the organization should have been made at a very early stage. Our findings suggest that the magnitude of work, given the resources available, was thoroughly understood. There seems to have been no concise plan for moving forward in the assigned roles; in consequence, precious years meant for planning and action were frittered away while participants bickered because the responsibilities for crucial activities were not clearly and concisely defined.

The sources and nature of the financing seriously limited Nigerian control and imposed penalties in the long run. The supply of technology and equipment and the erection of facilities were done by one firm (the Soviet firm, TPE). The execution of civil works, on the other hand, was mainly done by three contractors from Western countries. The financing of TPE and the Western contractors was through export credit. No major problems in this matter have arisen between TPE and ASCL. The delays and cost overruns have stemmed from the actions of the civil-works contractors, who have exercised total control over the rate of progress — without civil works, no erection of plants can take place. Therefore, the long gestation and cost overruns could be said to have arisen from four situations:

• the ambiguous nature of the terms and methods of payment set out in the contracts, particularly the anomalies in price-fluctuation formula (more than 2 years were lost in negotiations before an acceptable formula emerged);
• the lack of financial provisions for future escalations (no provisions were made, and so none was available when the need arose);
• the contractors' absolute control of the sources of financing; and
• the world economic situation (exchange-rate fluctuations added to the final cost of the project).

During execution, two major issues arose:

• the choice of self-administration (departmental execution) rather than a turnkey approach; and
• contracting of the civil works to three different firms.

Everything, from the market survey to the preparation of technical documents, had been done by either the Indian consultants or TPE. Although Nigerians can competently handle major aspects, there is dependence on consultants. This reliance on outsiders easily translates into a mistrust of the abilities of local staff. This malaise started from a lack of serious effort to organize a team for each of the critical areas around which the future plant could be built. One finds pockets of excellence scattered all over. They are not just badly organized — no organization is evident. The perception is that Nigerian engineers within the firm cannot possibly handle serious design work, despite evidence to the contrary. The need for teams is paramount for future modifications and expansions if the assimilation of technology is an objective. Umbrella units, like the "design bureau," could be used; this is a congregation of design engineering specialists who find solutions to technical problems.

When blessed by continuity, coherence, and excellence in composition, as well as by careful and systematic training in critical areas, teams have a chance of giving high returns. A team-building effort has, unfortunately, not been observed in the firm. It is essential, however, if the firm is ever to be free from reliance on foreign suppliers and consultants.

Because the planners were unable to appreciate the magnitude of the work, they failed to make the extra preparations needed to meet the challenges of the project. Grafting steel-making technology onto Nigeria — particularly at the scale attempted — represents a Gestalt shift. The complexity of the technology does not end when the hardware is installed; mastery of technology rests with the crucial supply of human elements. This dimension has not been given sufficient attention.

The experience with the civil works belied the idea that one can get the best value when contracts are shared among competing firms. When problems arose, ASCL was always forced to negotiate on three different fronts. This made the task arduous, unwieldy, and time consuming and may explain why negotiations to revise the price-fluctuation formula took such a long time. Financial arrangements that are acceptable to one firm may not suit another. Also, organizing the technological capabilities of the three firms added to the difficulties.

Nigerians may have been too optimistic in allocating to themselves such a magnitude of work, which in the long run affected the quality of planning. The quality of planning has come to haunt the project in several ways, and the "breathing space" provided by a long gestation was not usefully used in organizing to meet the future.

Additional causes of the long gestation include overruns before commissioning, as well as extra time spent trying to attain nominal plant capacity. It is not difficult to see the close correlation between the scheme of execution and the rate of progress — poor planning inevitably shows up in delays during project execution. What's more, the absence of teams with sufficient technical exposure will make an early start-up difficult. A realistic stocktaking of capabilities should precede the choice of implementation strategy. When the costs are added and compared with the benefits, one may find not only that time overruns have taken their toll in financial losses but that valuable skilled personnel have been lost during the long years of waiting.

As a guarantor of finance capital, the state was seriously disturbed by influences and powers external to it. This has been demonstrated clearly by the long delays in this project. The project's financing and Nigeria's inability to sustain foreign credit exposed the weakness and fragility of the state's financing power.

From the lessons learned, it is clear that the firm should try to ensure that it gets projects off the ground as speedily as possible and avoids harmful delays, and it should pursue the acquisition of technology so that it can rely on its own efforts.

Technology acquisition is a desirable social objective, but, to achieve it, firms must be able to shield themselves from the environment. They must garner sufficient technological capability to adapt strategically to environmental constraints. Aggressive pursuit of technical capacity and the attendant maturation forge for a firm a new image in the public eye. Then, a firm can chart technological growth that fulfils its objectives and its ideals. Clearly, to avoid falling prey to the lack of raw materials, inadequate infrastructure, and lack of local suppliers of parts, a firm must erect protective armour and then begin to set strategic goals.

Recommendations

1. Government should prepare adequately (through training) before attempting to execute a project with the ramifications of the Ajaokuta steel plant.
2. Decision-makers should make allowance for the acquisition of technology through two critical components: embodied knowledge (residing in humans and cumulatively acquired) and disembodied knowledge (exemplified in capital goods and all other physical manifestations of technology).
3. Efforts should be made to keep abreast of the state of the art in technology, innovative activities, and the strengths and shortcomings of suppliers and technical partners. This would ensure an optimum return on investment.
4. Steel-making projects and similar endeavours should be conceived as taking place within the context of the country. Practically all the capital goods for implementation of the Ajaokuta plant were imported. This state of affairs suggests that steel production was conceived of in isolation. The steel industry supplies the basic materials for the capital-goods sector, and the capital-goods sector supplies the machinery and equipment needed to erect plants for the steel industry. Industrial policy should articulate this symbiotic relationship, and a link should be actively promoted to ease the pain of project execution and promote indigenous manufacturing. It is difficult to see a truly indigenous technological effort taking root without a dynamic capital-goods sector.
5. The steel plant's facilities, such as the machine tools in ASCL's repair-shop complex, should be used to satisfy demands outside the steel plant. This would not only contribute to the economy but also help develop an indigenous technological capability and enhance adaptive abilities.
6. A rigorous inventory of resources should be taken of the raw materials needed in steel making. However, the technomanagerial abilities of the various institutions concerned with this would have to be upgraded first.
7. A study should examine the present status of the industrial sector and the adequacies or inadequacies of the institutional arrangements for finding, exploiting, and allocating resources and of those responsible for providing infrastructures.
8. Those making financial arrangements for projects should seriously consider the nature and source of funds. Indigenous equity financing and the encouragement of local capital financing would reduce the risk of the costly foreign-credit squeeze that has been the bane of this project's execution.
9. Rethinking is needed to correct the present anomalies in institutional coordination and to strengthen future organizations responsible for implementation.

References

• Dahlman, C.J. et al. 1978. From technological dependence to technological development; The case of the Usiminas steel plant in Brazil. Economic Commission for Latin America, Santiago, Chile. Studies in Technology 19B.
• Maxwell, P. 1982. Steel plant technological development in Latin America: A comparative study of the selection and upgrading of technology in plants in Argentine, Brazil, Columbia, Mexico and Peru. Inter-American Development Bank – Economic Commission for Latin America – United Nations Development Program – International Development Research Centre, Regional Programme of Studies on Scientific and Technological Development in Latin America, Buenos Aires, Argentina. IDB/ECLA/UNDP/IDRC Working Paper 55.
• NSDA (Nigerian Steel Development Authority). 1974. Preliminary project report. Project Review Committee, Lagos, Nigeria.
• UNIDO (United Nations Industrial Development Organization). 1983. Capital cost control of fertilizer plants in developing countries. UNIDO, Vienna, Austria. UNIDO/IS 422.
• ——— 1984. Manpower and training requirements in industry: Methodology with an application to the iron and steel sector. UNIDO, Vienna, Austria. Sectoral Working Series, No. 32.
• ——— 1986. Financial problems and the development of the iron and steel industry. UNIDO, Vienna, Austria. ID/WG 458/10.

Document(s) 21 de 29