15 Specifying, designing and optimising refrigeration systems In specifying refrigeration equipment the function of the equipment must be absolutely clear. Refrigeration equipment is always used to control tem- perature. Either the meat passing through the process is to be maintained at its initial temperature, for example as in a refrigerated store or a packing operation, or the temperature of the meat is to be reduced, for example in a blast freezer. These two functions require very different equipment. If a room is to serve several functions then each function must be clearly iden- tified. The optimum conditions needed for that function must be evaluated and a clear compromise between the conflicting uses made. The result will inevitably be a room that does not perform any function completely effectively. There are three stages in obtaining a refrigeration plant. The first stage is determining the process specification, the second stage is drawing up the engineering specification, i.e. turning processing conditions into terms which a refrigeration engineer can understand, independent of the food process, and the third and final stage is the procurement of the plant. 15.1 Process specification Poor design in existing chillers/freezers is due to a mismatch between what the room was originally designed to do and how it is actually used. The first task in designing such a plant is therefore the preparation of a clear speci- fication by the user of how the room will be used. In preparing this speci- fication the user should consult all parties concerned.These may be officials enforcing legislation, customers, other departments within the company and engineering consultants or contractors, although the ultimate decisions taken in forming this specification are the user’s alone. 15.1.1 Throughput The throughput must be specified in terms of the species to be handled and whether they are split, whole, quartered, primal joints, and so on. If more than one species or type of cut is to be processed then separate specifica- tions must be made for each product. The range and average weight and fatness of each product should also be specified. For example, large car- casses can take twice as long to chill as small carcasses under the same con- ditions, so it is important to be realistic in deciding on the weight range. To say that all types and weights of animals slaughtered are to go through one chiller or freezer will inevitably mean that compromises must be made in the design stages which will lead to an inadequate system. A throughput profile is needed. Few meat plants slaughter the same number and weight of animals on each day of the week and therefore the average throughput is not adequate in the specification. The maximum capacity must be catered for and the chiller/freezer should also be designed to chill/freeze carcasses adequately and economically at all other throughputs. 15.1.2 Temperature requirements The range of temperature requirements for each product must also be clearly stated. In deciding what this is or these are, several requirements, often conflicting, must be considered. First of all, what legislative require- ments are there, for instance the EEC requirement at 7 °C? What customer requirements are there? These may be your existing customers or they may be future customers who you are hoping to attract. What standard do you yourself have? Some companies sell a quality product under their own brand name, which should include a cooling specification. Finally it must be decided to what extent the above standards may be allowed to slip. The reason for this will become apparent later. Almost everyone in the meat industry allows their standards to slip to some extent; those that get caught and are called to task for this lose orders or have their production disrupted. These firms have turned their back on this problem and not dealt with it consciously and clearly.There are others who know well to what extent they can push the inspectors, or their customers and ensure that they stay within accepted limits. 15.1.3 Weight loss If it is intended to save weight from the meat both during chilling/freezing and storage, it is useful to quantify at an early stage how much extra money can be spent to save a given amount of weight. 304 Meat refrigeration 15.1.4 Future use All the information collected so far, and the decisions taken, will be from existing production. Another question that needs to be asked is, will there be any changes in the use of the chiller/freezer in the future? In practice the answer to this question must always be ‘yes’. Looking back into the past, no meat processor has remained static and within the foreseeable life of a chiller/freezer, anything between 10 and 50 years (judging by present chiller/freezer population), can any changes be envisaged and can these be quantified in as much detail as possible? It is still not possible at this stage in the design to finalise the factory layout and operation. However, some estimate of how the factory will be operated, how it will be laid out, the size of chiller/freezer needed, and so on must be made at this point. This must be kept flexible until the engineering specification has been formulated (see later). It is common practice for the factory layout and operation to be decided in advance of writing the chilling/freezing specification. Lack of flexibility in changing these is often responsible for poor chiller performance once the factory is completed. 15.1.5 Plant layout Chilling or freezing is one operation in a sequence of operations. It influ- ences the whole system and interacts with it. An idea must be obtained of how the room is to be loaded, unloaded and cleaned, and these operations must always be intimately involved with those of the slaughter line, the sales team, the cutting and boning room, and the loading bay. Questions that need to be addressed include ‘where will meat be sorted for orders?’ and ‘where will meat not sold be stored until a future date?’ There is often a conflict of interests within a chiller/freezer. In practice the chiller/freezer is often used as a marshalling yard for sorting orders and as a place for storing car- casses that have not been sold. If it is intended that either of these opera- tions are to take place in the chiller/freezer the design must be made much more flexible in order to cover the conditions needed in a marshalling area or a refrigerated store. Meat must be loaded into and out of the freezer or chiller and the process may be continuous, batch or semicontinuous. In the case of batch and semi- continuous processes, holding areas will be required at the beginning and end of the process in order to even out flows of material from adjacent processes. The time available for the process will be in part dictated by the space available; a slow process will take more space than a fast process, for a given throughput. It may also be dictated by commercial constraints, such as the delivery of ‘1-day-old’ meat to distribution outlets. The above specifications will dictate the processing conditions. Most processes use air as a processing medium and its temperature, velocity and relative humidity are all usually critical to the process. The processes may Specifying, designing and optimising refrigeration systems 305 be in a single stage, in which case steady values will be specified, or they may be time dependent, as in a multi-stage process. In choosing the process conditions there will be an interaction with the earlier specified constraints. Some compromise may be needed, adjusting the time available for the process in order to obtain an optimal solution. Once the process conditions are fixed and the throughput and materials specified, the product load will also be fixed although this may not always be known. Where design data exist, they should be utilised to specify the product load. Other refrigeration loads also need to be specified. Many of these, such as infiltration through openings, the use of lights, machinery and people working in the refrigerated space, are all under the control of the user and must be specified so that the heat load given off by them can be incorpo- rated in the final design. Ideally, all the loads should then be summed together on a time basis to produce a load profile. If the refrigeration process is to be incorporated with all other processes within a plant, in order to achieve an economic solution, the load profile is important. The ambient design conditions must be specified. These refer to the temperatures adjacent to the refrigerated equipment and the temperatures of the ambient surroundings to which heat will ultimately be rejected. In stand-alone refrigerated processes this will often be the wet and dry bulb temperatures of the outside air. If the process is to be integrated with heat reclamation then the temperature of the heat sinks must be specified. Finally, the defrost regime should also be specified. There are times in any process where it is critical that a defrost does not take place and that the coil is cleared of frost before commencing this part of the process. The above requirements should all be specified by the end-user. It is common practice throughout European industry to leave much of this specification to refrigeration contractors or engineering specialists. Often they are in a position to give good advice on this. However, since all the above are outside their control, the final decision should always be taken by the end-user, using their knowledge of how well they can control their overall process. 15.2 Engineering specification The aim of drawing up an engineering specification is to turn the process- ing conditions into a specification that any refrigeration engineer can then construct and deliver without knowledge of the meat process involved. If the first part of the process specification has been completed, the engi- neering specification will be largely in place. It consists of the environmen- tal conditions within the refrigerated enclosure, air temperature, air velocity and humidity (the way the air will move within the refrigerated enclosure), the size of the equipment, the refrigeration load profile, the ambient design conditions and the defrost requirements. The final phase of the engineering 306 Meat refrigeration specification should be drawing up a schedule for testing the engineering specification prior to handing over the equipment. This test will be in engi- neering and not product terms. During this process the user must play an active part because a number of the decisions taken in this stage will affect other aspects of the opera- tion. The specification produced should be the document that forms the basis for quotations, and finally the contract between the user and the contractor must be stated in terms that are objectively measurable once the chiller is completed. Arguments often ensue between contractors and their clients from an unclear, ambiguous or unenforceable specifica- tion. Such lack of clarity is often expensive to all parties and should be avoided. 15.2.1 Environmental conditions The first step in this process is iterative and is shown symbolically in Fig. 15.1. First, a full range of time, temperature and air velocity options must be assembled for each cooling specification covering the complete range of each product.The list should also include future cooling specifications. Each must then be evaluated against the factory operation. For example, using a particular temperature and air speed around one product may give a chill- ing time to meet the temperature requirements already laid down of 18h. If the factory operation calls for maximum chilling time of only 12 h then clearly the temperature/time combinations currently under review will not Specifying, designing and optimising refrigeration systems 307 Temp./vel./time options for a product range Fit plant operation? NO NO NO YES YES Is there another option? Alter standards or factory operation YES Is this best option? Final temp., time, RH & times for product range Fig. 15.1 Flow diagram for a selection of the environmental conditions. fit. Therefore another option should be selected and the process repeated. If there are no more options available there are only two alternatives: either standards must be lowered, recognising in doing so that cooling specifica- tions will not be met, or the factory operation must be altered. Having found a temperature/velocity/time option that fits in with the plant operation the next question is, ‘is this the best option?’ All possible options must be evaluated in order to ensure that the optimum is obtained. From this a final set of times, temperatures, air speeds and relative humidi- ties will be obtained. In a chiller/freezer intended for several uses with product going to different customers this list can be quite long and if future uses are also included it can be longer still. 15.2.2 Room size Using throughput information from the user specification and the chill- ing/freezing times now worked out, the size of the room can be determined. To achieve this, the operation of the whole abattoir may have to be changed and also the flow of carcasses to and from the chiller/freezer, the position of doors, and so on. If the size and position of the room has been rigidly fixed before this stage, the cooling times determined above will not be met. 15.2.3 Refrigeration loads Refrigeration load calculations can now be performed, leading to a load profile for the room. Table 15.1 shows a typical load profile for an ‘imaginary’ beef chiller.This chiller is to have two uses: first, to chill beef in a single stage process, and second, to store previously chilled beef sides. The product loads have been worked out for three separate conditions (in practice more may well be needed), the peak load, which will occur at the end of the loading period, the average load during chilling and the load on the room when it is used for storing previously cooled sides of beef. The peak product load can be obtained from data provided in other chapters of this book. This load is very high and occurs for only a short period of time. The average product load is then calculated, based on the amount of heat to be extracted over the entire chilling period divided by the time available for chilling. Finally, when the room is used as a chill store there will be no product load in the room. The infiltration load is the next most important feature yet there is little published information. The best is possibly in the ASHRAE guide. When loading a beef chiller the doors are invariably left open for long periods allowing a fully established air flow to take place to and from the room either from gravity through a single door or by a through flow of air if more than one door is open. Designers often decide that the door will only be 308 Meat refrigeration open for short periods and that fully established airflow will never occur. However, most chiller doors stand open for ca. 8h in every 24h and they are certainly open while the chiller is being loaded. This infiltration load occurs at the same time as the peak product load, i.e. immediately at the end of the loading period, and it is therefore important that the infiltration load is added to the product load to find the maximum loading on the chiller. Experimental data obtained by FRPERC have shown that the ASHRAE calculations are approximately correct and are therefore rec- ommended. They are often of an order of magnitude higher than figures used by designers of chill rooms. The infiltration load during chilling, when the average product load occurs, is much smaller because the door is nor- mally closed or only opened for short intervals. It is therefore more accept- able to use quite small infiltration loads during this time to allow for air exchange through faulty door seals or for short door openings only. When the room is being used as a store, particularly if it has been unloaded, there may be periods when the infiltration load rises to maximum levels again and this has been shown in Table 15.1. The fabric load is shown next in Table 15.1 and is insignificant compared to the previous two loadings. The same applies to the heat load imposed by people in the chiller and by lighting. Unfortunately both these loads are normally concurrent with the peak product and infiltration loads and must therefore be added to these to calculate the total peak load. The evaporator fans also produce quite high loads. At this point in the design an approximate figure for evaporator fan power must be used but when the final design is completed and more accurate data are available, this must be substituted and the calculations reworked. In the current Specifying, designing and optimising refrigeration systems 309 Table 15.1 Refrigeration loads for a beef chiller Refrigeration load (kW) Peak Average Store Product load 40 16.8 nil Infiltration 17 0.5 0.5 (<17) Fabric 2 2 2 People 0.25 nil nil Lighting 1.5 nil nil Evaporator fans 6 6 6 Contingency – 2.5 (10%) – Defrost – (<12) (<12) 18 h running – – Totals 66.8 27.8 8.5 (<25) ‘Peak’ refers to the maximum load at the end of loading, ‘average’ to the load during chilling when the product is at its average value, and ‘store’ to the loads present when the chiller is used to store previously cooled sides. example the evaporator fans are left running continuously and therefore the load is the same under peak cooling conditions as when the room is used as a store. A contingency or safety factor is often added to the above calculations, to allow for errors. In the table of loads this has been added to the average figures, but left out when calculating the peak load. This is because the peak load occurs for only short times and errors in it have less effect than at other times. Some designers add on an ‘allowance for 18 h running’. This needs some explanation. In many cooling applications the loads on the refrigeration plant are fairly constant over a day. If the refrigeration plant were installed to meet these loads then it would be running for almost 24 h a day in order to meet them. This is felt to be ‘bad’ for the refrigeration plant and it is considered desirable that the plant should have rest periods in between run times and therefore an allowance is added to ensure this. Such an allowance is irrelevant in a batch chilling process where a peak load rapidly declines. It is also often only used as a euphemism for a (quite large) contingency. When a room is defrosted, cooling stops and some heat is added. When the refrigeration plant reverts to cooling there is an additional heat load on the room. It is a wise precaution to ensure that defrosts do not occur at the same time that peak cooling is required and therefore no further allowance needs to be added to the peak heat loads. However, defrosts may be needed during cooling and when the room is used as a store and therefore the extra cooling has been shown in Table 15.1, but not added into the totals. In this example of an ‘imaginary’ chiller the total heat load is over twice the calculated average heat load. Thus if the refrigeration plant was sized to meet only the average load it would only have half the required capac- ity. Another point to notice is that the load on the room, when used as a store, even when the outside temperatures are very high, is very small com- pared to both the peak and average loads and is for the most part due to the evaporator fans running continuously. These points will be considered in more detail below. The load profile can now be plotted and, for the hypothetical example already discussed, is shown in Fig. 15.2. At the start of the plot the room is running with no product load and with the doors closed. The load then increases when the doors are opened and the room is washed out or pos- sibly unloaded. Warm carcasses are then loaded into the room and the load rapidly reaches the peak product load that occurs at the end of the loading period. Thereafter, the doors are closed and the load rapidly declines. At the end of the chilling cycle, the doors are again opened to remove the car- casses and the infiltration load so caused increases. During the chilling period additional peaks occur after defrosts. Note that the plant only runs at its full or peak load for less than 4% of the time. It runs at or above half load for 25% of the time and for over 70% of the time runs at less than a quarter of its design load. 310 Meat refrigeration 15.2.4 Refrigeration plant capacity The capacity of the refrigeration plant must now be decided. Will the peak heat load be met? If so the planned chilling times and the specification agreed will be achieved. If not, the desired schedule will not be attained, but there will be a saving in the capital costs of the refrigeration plant and more economical running will be achieved since the plant will be running at less than quarter capacity for over 71% of the time. Large refrigeration plants working at low loads are very inefficient and therefore very costly to run. If a smaller plant is used, it will have a smaller turn-down ratio and its efficiency at part load, the majority of time that it will run, will be higher, therefore saving operating costs. There are some possible solutions to the designer’s dilemma. If refrig- eration capacity is demanded elsewhere in the plant, but at different times, the provision of a central plant serving both facilities can make use of this diversity. It is therefore important at this stage to look at the refrigeration load profile for the entire plant – there may be blast freezers which are only operated well after the time that the peak chilling load has passed and by careful design a refrigeration plant may be installed and shared between both facilities. However, this only applies to a part of the refrigeration plant, the compressors and condensers, and not the evaporators. Another solution Specifying, designing and optimising refrigeration systems 311 Load Chillier Door open. unload/ washoown Unloading Load profile 80 60 40 20 0 10020304048 Time (h) Fig. 15.2 An example of a refrigeration load profile for a beef chiller. is to install a plant that is smaller than needed and recalculate the extended cooling times that will occur, still meeting the chilling/freezing specification. When the refrigeration load increases above that at which the refrigeration plant can extract air, temperature in the chiller/freezer will rise. As the air temperature rises two things happen: the product load is reduced and the capacity of the refrigeration plant will increase. The effect of these two changes is that after a time a balance is achieved when the load arising from the carcasses is extracted by the plant and cooling can then continue in the normal way until the temperature is reduced back down to its original level. If data were available on cooling during pull-down it would be possible to recalculate the extended cooling periods, check whether these fitted with the other user requirements and still install a refrigeration plant which would meet an agreed specification. Another option is to spread the loading time of the chiller/freezer over a longer period, and so reduce the peak product loads. However, this normally causes disruption in the planned operation of the abattoir, with decreasing productivity, and is therefore rarely used. Whatever decision is taken, the peak product load that the refrigeration plant is expected to accommodate should be clearly stated in the agreed engineering specification and a load profile should also be given to ensure that the refrigeration designer provides a plant that will run efficiently over the entire product load range. 15.2.5 Relative humidity The relative humidity in the chill room must also be defined in the engi- neering specification. When the chiller is empty and the latent heat load negligible, relative humidity depends upon the evaporation temperature. The wet bulb temperature will slowly be reduced until it approaches evapo- ration temperature, at which point no more water will be extracted from the room and the humidity will remain stable. As the latent heat load in the room increases from the loading of warm wet carcasses, the amount of water vapour evaporated increases the relative humidity in the room. The relationship between latent heat load and relative humidity depends upon two factors: the design of the evaporator coil and plant cycling. Only an infi- nitely deep coil will produce discharged air with a wet bulb temperature equal to the evaporator temperature. During refrigeration plant off-cycles, no water is extracted from the air and the relative humidity will rise, only to be pulled down again when the refrigeration plant is switched back on. Therefore, to obtain a high relative humidity in the room to reduce weight loss during the latter parts of chilling and storage, a high evaporation tem- perature and a large coil area are needed. This ensures that the refrigera- tion plant runs for only very short periods of time and when it is running it only extracts the minimum amount of water from the air. Since humidity is normally only important in the latter stages of chilling, 312 Meat refrigeration when the load on the refrigeration plant is small, relative humidities greater than 90% can be specified. The engineering specification should specify the relative humidity under full sensible heat load, i.e. the lowest hypothetical relative humidity that can be obtained, and also under part load conditions, when advantage can be taken of the reduction in load to raise evaporation temperatures in the space. 15.2.6 Ambient design conditions The conditions in the air outside the chiller/freezer must also be defined in the engineering specification. Both the infiltration and fabric loads are dependent on the outside temperature which therefore has an important effect on the capacity of the refrigeration plant. Ambient temperature also affects the capacity of the refrigeration plant because heat must be rejected above this temperature via a cooling tower or condenser. If it is intended that the room should function under all possible ambient conditions, very high ambient wet and dry bulb temperatures must be specified. However, these normally occur only during exceptional circumstances and, only briefly at or soon after midday. For design purposes, temperatures that are not exceeded for more than 2.5% of the total time in the year are normally acceptable and often a figure of 5% is used. Both wet and dry bulb tem- peratures should be specified, giving the option of using an evaporative- type of condenser or cooling tower for heat rejection to the atmosphere, which leads to a more efficient and smaller cooling plant. 15.2.7 Defrosts The occurrence of defrosts should also be specified to avoid peak periods while still ensuring that during these peak periods the evaporator is clear of ice. It is normal in abattoirs to defrost the evaporators and the chillers at 6 hourly intervals. Although this is often desirable immediately after the peak latent heat load has been removed from the room, it is unnecessary during the later stages of cooling and when the room is used as a store. Limiting defrosts can reduce both energy consumption and weight loss during subsequent use and therefore should be included in the engineering specification. 15.2.8 Engineering design summary The engineering specification should therefore include each of the items shown below: ? chiller, freezer air temperature, air speed and relative humidity for each product specification (covering complete range) and the time that each of these periods will be operating, Specifying, designing and optimising refrigeration systems 313 ? the ambient air temperature, both wet and dry bulb, ? the load – peak, average and store, ? infiltration load, i.e. the number of doors and the time they will remain open, under what circumstances and conditions, ? evaporator and condenser temperatures. All the conditions laid down in the engineering specification can be mea- sured and therefore do not depend upon variation in usage or even abuse of the chiller and should therefore form the basis of a contract. 15.3 Procurement The engineering specification should be sent out to tender. If tenders have been selected for the quality of their equipment and all accept the tender conditions and say they can meet the design and test conditions specified, the lowest tender would normally be chosen. The contractor is normally responsible for the detailed engineering design, construction and commis- sioning and the only need is to check that this work is carried out in a pro- fessional way. The first responsibility of the contractor is to carry out the acceptance tests.These test the performance of the refrigeration equipment in terms of the engineering specification, and the plant should not be accepted until satisfactory tests have been carried out. The plant can then be handed over and training given to the plant operators in the correct use of the refrigeration equipment. The plant then needs to be commissioned by the factory personnel, systematically increasing throughput until process tests can be carried out. These ensure that the original process specification actually achieves the intended results in terms of temperatures, through- puts and yield. 15.3.1 Plant design Once the engineering specification has been written and agreed, the plant design is relatively straightforward. The difficult decisions have already been taken. The details of refrigeration plant design have been laid out in many other textbooks and will not be repeated here. However, one problem area is specific to carcass chillers and this is the selection of evaporators. 15.3.1.1 Evaporators An evaporator has two functions. The first function is to remove heat and moisture from the air in order to maintain the correct design conditions. The second function is to move air around the room. Standard evaporators are designed for the first function, but the second is largely ignored and some evaporator manufacturers do not even publish information on how 314 Meat refrigeration their evaporators may be used for this purpose. This is partly because air movement in the room not only depends on the evaporator design but also depends on the shape of the room, position of the beams, and so on. Some general points follow. Air movement is controlled by blowing air and not sucking; this means that it will move under the effects of dynamic or static pressure. A moving jet of air will not go on moving forward indefinitely but will be slowed down by friction against solid surfaces and by friction and entrainment of static air adjacent to it. The throw of the jet is the maximum distance that the air will move until it slows down to a specific velocity. If the air is thrown under a ceiling or along another solid surface, the coriander effect will increase the length of the flow, whereas if the air is thrown out into a void with fric- tion on all sides, mixing and entrainment will slow the jet in a shortened distance. Any beams, light fittings or solid objects in the path of the jet will serve to deflect it and turn it in another direction and therefore have a pro- nounced effect upon the air movement within the space. Air can be distributed in a space above the rails, possibly with a false ceiling or via plenums or ducts. Estimates of air movement in a chiller are sometimes based on the cross-sectional area and the volume of air passing through it. Although this gives an average velocity over that cross-sectional area it is often very deceptive since this velocity rarely occurs around the carcasses. Such calculations are more useful in designing blast freezers when a horizontal flow of air moves through a tunnel. They may sometimes be used in prechillers which function as blast chillers, similar to blast freezers with high air velocity and a large refrigeration load and consequently a large column of air movement. In Fig. 15.3 several ways of distributing air from the evaporators around a chiller are shown. Each uses the void above the rail for primary distribu- tion. As air is blown under the ceiling it is drawn across by the coriander effect towards the far wall. On the way air is entrained from around the carcasses and drawn into the jet until the jet expands and starts hitting the rails. This deflects a proportion of air down, around and through the car- casses. If the velocity of the jet is more than the maximum velocity required around the carcasses, by the time it reaches the far wall (throw), or another jet, a down draught will occur at that point, moving to the floor and then in turn be deflected back across the floor. Some problems that occur arise from a conflict between the two func- tions of the evaporator. If it is selected to meet the peak product load, the air velocities created can be excessive, i.e. the throw will be greater than the distance from the evaporator face to a wall or jet facing it. Sometimes the situation is reversed, i.e. the throw is too small. There are several possible solutions: ? the evaporator selection can be changed to alter the discharge velocity, ? the fans can be altered from draw through to blow through or vice versa, Specifying, designing and optimising refrigeration systems 315 ? the discharge angle of the jet can be altered, ? the position of the evaporator can be altered. In the final design a combination of the above can be used. However, in the authors’ opinion, although there is considerable infor- mation relating to air movement there is still insufficient information avail- able to design a uniform air velocity around all carcasses and it must be recognised there will be a range of air velocities in the room and hence a range of cooling conditions. The largest refrigeration load during carcass storage and in the latter stages of cooling is that of the evaporator fans. These are selected to extract the peak-cooling load where high velocities are needed. There are several 316 Meat refrigeration PlanPlan AA AA A A NN N N N N N N N N A A A A AA Plan Fig. 15.3 Examples of air movement produced by different evaporator positions. options available to the designer to reduce excessively high air velocities during the second stage of chilling and during storage. The first is to use two-speed motors; another is to switch off a number of fans on each evaporator unit or even a number of the units. Although this will cause an uneven air distribution within the store, it has often proved adequate to maintain temperatures throughout the room during storage. Often the con- flict is so great that such solutions are not possible. In these circumstances it would probably be better to design a separate refrigeration system for use when the chiller is used as a store. For instance, one refrigeration plant is designed for chilling a separate plant with a sock diffuser system used for storage, giving ideal velocities during both functions. The smaller refriger- ation plant installed for storage will operate much nearer its maximum load for long periods and hence be far more efficient than the much larger chill- ing plant operated under the same circumstances. 15.4 Optimisation This covers both existing and new processes. The procedure is similar in both cases with the difference that there are few constraints on new processes but there are less design data available. The first step will be definition, the next will be looking at improvements and this will be sub- divided into improving process conditions and equipment and improving existing equipment and its use. 15.4.1 Process definition The first step in the process definition is to define, in objective terms, what the process consists of. As in specifying equipment, the user must quantify the throughput, the initial and final temperatures, and any change in yield and/or temperature that is desired in the process. The next step is to measure what has actually been achieved and what process flows are needed to achieve it. The latter will consist, for example, of the amounts of water, energy, labour and so on that are required by the process. The next step is to identify what limitations or constraints there are on the process. In existing processes this will often be the time available for the process, the space available, possibly limitations on the amount of power available and financial constraints. The final step is to compare the existing process with an ideal process as a measure of the efficiency improvement. If process design data are available they can be used to optimise the process within the constraints already listed to see how far it can be moved from its present performance towards the ideal. Often process design data are not available. For new processes these must be obtained through the use of mathematical models or pilot plant trials. In existing processes, the existing plant can be used to see the effect of changes of each variable. Specifying, designing and optimising refrigeration systems 317 Initially, only two values of each variable will be used. A theoretical analysis of the process will often indicate whether there will be an interac- tion between process variables. If there are, each level of a variable should be tested against each level of all other variables.Analysis of the results will indicate which variables are significant and in which direction they need to be moved. The significant variables can then be investigated more fully, often using a stepwise process, adjusting each variable slightly until the process no longer improves, then changing to the next variable and repeat- ing the process. Final values should be evaluated in a separate trial in order to obtain a benchmark for future improvement. Often the optimum process conditions are already known but measurements in the existing process show that these are not being achieved. A targeting and monitoring system of quality control is, therefore, required. The process variables should be monitored, the results over a sig- nificant period analysed and targets set. The results of the monitoring and the targets should be clearly indicated to the operators and maintenance staff. Subsequent monitoring usually shows a marked improvement in per- formance. If problems still persist then it may be that additional training of staff is required in order to improve their performance. 15.4.1.1 An example of the system in use Customer complaints were being received by a meat processing factory that the temperature of meat being delivered was warmer than required. This prompted the factory to investigate the performance of their blast freezing operation. Initially, only the air temperatures in the blast freezer were monitored. This indicated that during the night shift staff were leaving the doors open for long periods, causing large increases in temperature. It was further realised that the temperature increased progressively during defrosts. The maintenance staff became aware that there was a serious problem with the defrosting, which was slowly building up to a crisis point. This prompted them, on their own initiative, to clear the coils completely of ice and to instigate more frequent defrosts. However, temperatures in the frozen store did not improve and it was realised that more frost was forming on the coils than had been allowed for in the initial design. The monitoring was increased to measure the defrost water collected from the coils. The main source of frost on the coils was from infiltration through the doors to the frozen chamber.This indicated that the doors were used far more frequently than allowed for in the original design. In order to optimise the performance of the system within the constraints of the factory, a monitoring system was placed on the doors. In addition a new loading and unloading system was introduced to limit the number of door openings and the staff instructed in its use. As a result, the plant was able to deliver meat to its customers at the correct temperature. After a while, the monitoring system was allowed to lapse and with the absence of feed- 318 Meat refrigeration back the problem returned and customer complaints were again received. The solution was to reinstate the monitoring of the cold room doors, of the cold store air temperature and the amount of defrost water collected and to feed the results back on a regular basis to the operating staff. Because they had already been trained in the correct use of the store this was suffi- cient to correct the problem. The point of this example is that defrosts are often ineffective. If the defrosts do not occur frequently enough, then the evaporator can become sufficiently blocked to impair performance. This can cause a build up of ice which it is not possible to remove within the normal defrost period. If the defrost period does not last long enough in order to remove the ice completely from the evaporator, then the same problem will occur again. Normal settings for defrost in such a room would be half an hour every 6 or 12 h. It is possible that in the past the number and extent of defrosts had been insufficient in order to remove ice from the coils. 15.4.1.2 Action needed to maintain the performance of the room in the future The user of the cold room is instructed by the contractor on how to inspect the evaporator for build up of ice. When this is first observed the user is instructed on how to force extra defrosts in order to remove the ice before it becomes a problem and impairs the performance of the temperature control of the room. The contractor and user should agree on the maximum number and duration of door openings and the user should ensure that these opening times are not exceeded. The user of the cold store should plan the move- ments of frozen material in and out of the room in order to minimise the time that the cold store remains open. The roller shutter door at the front of the unit does not need to extend to the full height of the door opening. It could be stopped short at the top and louvers fitted in the remaining space above the door. The discharge condensing unit can then be ducted to dis- charge air directly outside, drawing fresh air back in over the unit. Thus, the air temperatures around the condensing unit can be considerably reduced. Should the amount of time which it is considered necessary to have the cold store open for the movement of frozen goods prove to be consider- ably greater than that allowed for in the design, then the entrance to the cold store should be modified in order to reduce the amount of warm, moist air entering the room while the door is open. This could be effected by con- structing a lobby or vestibule around the cold store, which is refrigerated to chill room conditions. Frozen material should be brought through into this chilled area first, then the door between the chilled area and the shop closed. The door between the cold store can then be opened and the mate- rial loaded from chill room into the cold store. This will reduce the amount of water vapour entering the cold store by a factor of ca. 70%. Specifying, designing and optimising refrigeration systems 319 15.5 Conclusions The performance of refrigeration systems is a major source of conflict between users and refrigeration contractors. The adoption of the approach outlined in this chapter should avoid these conflicts. If performance prob- lems occur then the approach will clearly identify which partner is respon- sible for sorting out the problem. Users must accept responsibility for the process specification. They must clearly identify what they wish to achieve and take into account their expan- sion plans. Outside input may help them clarify their requirements and options but the final decision is theirs. From a clearly defined process specification, an engineering specification can be written that defines the requirements in terms of the conditions that have to be produced within the refrigerated enclosure.This will define space limitations, for example, the tests to be carried out before acceptance and any monitoring/control instrumentation. The process specification and its development into an engineering speci- fication are the critical steps in obtaining a system that works. The rest of the procedure should follow automatically. However, the cost factor should not be forgotten. Often the initial quotations are outside the user’s budget and during discussions cost savings are agreed. All too often, there is an implied change to the engineering specification and consequently to the process specification. Unless this is formally recognised and the specifica- tions amended to take into account the change, a source of conflict in the future has been established. Once the plant has been constructed and commissioned, routine moni- toring and action when performance changes are the final key tasks in the maintenance of an optimum refrigeration system. 320 Meat refrigeration