3.2 DESARROLLO DE LA PROPUESTA
3.2.17 REQUERIMIENTOS Y CONSIDERACIONES OPERACIONALES
3.2.17.3 REQUERIMIENTOS OPERATIVOS
The next actor upstream from the millers is the Food Reserve Agency (FRA), an overview of the corresponding model structure is visible in figure 10.
Figure 10: FRA structure
FRA has different kinds of storages in many locations around Zambia. Farmers first bring their maize to the nearest local FRA sales point, which normally consists of temporary storage facilities, such as slabs. These facilities are usually not well protected and thus exhibit high loss ratios. Maize is then moved from there to satellite storages in regional
centres, which feature more elaborate permanent storage facilities, such as sheds and silos. Sheds, however, still have quite high loss ratios in the medium term and only silos really keep maize largely unaffected by pests, moisture and fungi. (Bou Schreiber, 2015)
I therefore assume that FRA puts as much maize into their silos as their silo capacity allows. Once these are full, the next best storage option is to put the maize into the sheds. And only if these are also already full (something that only occurred in the bumper harvest years 2010-‐14 when FRA bought very large amounts of maize), maize is left in the local slab structures.
Building a model structure that could adequately measure the maize loss in FRA storages proved to be one of the trickiest parts of the work. After experimenting with different solutions it became increasingly clear that I could not work with the traditional way of measuring residence times (dividing a stock by its outflows), because the strong seasonality of my model lead to extreme fluctuations in the residence time that did not yield any realistic values. I thus adapted the age measurement structure from Bou Schreiber (2015) which proved to work well despite the seasonality in my model. It works the following way: when new maize is purchased, the current time (in months) is multiplied by the purchase volume (in weight), and the thereby “age-‐tagged” maize volume is accumulating in the co-‐flow structures visible in figure 10 above the value chain flows. Thus, dividing the maize currently on storage by the accumulated age-‐weight stock in the co-‐flow returns the average time when the maize on storage came into the stock (average entry time).
As maize is sold, old and new maize are assumed to be mixed for sale, so that on average, the maize sold (or moved to better storage facilities, in case of the first FRA stock) has the average age of all the maize in store. This procedure, as opposed to e.g. selling only the oldest maize first, is followed because the longer maize is on storage, the more its quality is adversely affected by moisture, fungi etc. – and the stronger quality is affected, the lower is the price for which the maize can be sold. Therefore, old maize is mixed with newer maize
in order to be able to sell it for a good price. For more details on this, see Bou Schreiber (2015). The outflow from the age-‐weight co-‐flow is thus not multiplied by the current time, but by the average age (Average entry time) of the maize in stock.
Now, to measure the average time that the maize in stock has actually spent on storage (the “relative age”), one simply has to take the difference between the current time and the average entry time. The relative age (i.e. time spent on storage) of the stock then determines the loss ratio, which is exponentially rising with relative age to reflect the exponential effects of spreading moisture, fungi and pest populations. Multiplying the loss ratio with the stock then determines the actual loss outflow.
Since Bou Schreiber's (2015) original structure was intended to measure the losses over one annual cycle, the outflow has to be divided by 12 months in order to transform the time horizon and return the correct loss values. In the case of the permanent storage structure, the combined loss ratio also depends on which ratio of the maize is in silos and which ratio is in sheds, since the two storage types exhibit distinctly different loss ratios (cf. appendix C.3).
FRA then sells its maize according to domestic demand to contracted millers, but not to retailers or consumers (N. M. Mason & Myers, 2011). This is important since it means that maize once purchased by FRA is effectively locked up in the formal value chain and cannot be supplied to informal consumers any more. If there is still excess maize left in FRA’s storages at the end of the marketing year and it cannot be sold domestically, FRA exports the amount exceeding its desired security stocks. For more details on these exports, see section 4.3.4.
Furthermore, I assume that FRA wants to keep a certain share of security stocks to fulfil their original mandate, which is stabilizing the maize price with strategic maize reserves. However, I included a switch structure that sets the desired reserves to zero when we have a year with a structural maize deficit. This reflects that FRA, whose mandate is secure the
maize supply for the population, would not keep stocks locked up while domestic consumers demand more maize to fulfil their basic needs.