Measure and control and location of CCPs is the second principle in HACCP design and is really the heart of the HACCP plan. A CCP is a step at which control can be applied and is essential to prevent or eliminate a food safety hazard, or to reduce it to an acceptable level . For every significant hazard identified, one or more CCPs must be designated to control or eliminate the hazard. For example, the thermal process given to canned vegetables at the retort step would be designated as a CCP in the low-acid foods industry, while proper acidification or pH adjustment in the brine kettle step would be the CCP in the acidified foods industry. Sometimes it is not possible to eliminate a potential microbiological hazard, only to minimize it to an acceptable level. In the fresh-cut produce industry, proper water chlorination at levels > 1 ppm free available chlorine in flumes and dip tanks becomes the CCP .
Unfortunately, there is no simple, clear-cut answer to the question of how many CCPs a HACCP plan may need and where should they be located. It depends on plant layout and design, the product being produced, the ingredients used, equipment age and condition, processing methods employed, and, especially, the effectiveness of the prerequisite programs implemented. Often an SSOP or SOP can be incorporated to control a hazard rather than a CCP. To keep HACCP programs plant-friendly and sustainable, Bernard recommends that the number of CCPs be kept to a minimum, and none should be redundant . Redundancy will also add to the cost of record keeping. Experience has shown that HACCP plans that are unnecessarily cumbersome will likely be the ones that fail.
However, many points in a flow diagram not identified as CCPs may be considered control points. A control point (CP) is any step at which biological, physical, or chemical factors can be controlled . Many types of control points can exist in fruit and vegetable operations, including those that address quality control (color, flavor, texture), sanitary control (SSOPs, GMPs), maintenance (calibration of equipment), and process control (fill weights, seal closures).
Pinpointing the right CCPs is the most crucial and problematic aspect of an effective HACCP program . Therefore, a common strategy to facilitate the proper identification of CCPs is to use the CCP decision tree (Figure 15.2) [7,41]. The decision tree consists of four questions that are asked for each process step for which hazards have been identified during hazard analysis. The answer to each question will direct the process of elimination and ultimately lead to a decision as to whether a CCP or CP is required at that step. A benefit of the CCP decision tree is that it forces and facilitates HACCP team discussion and teamwork and ensures a consistent approach to every hazard at each step . However, as pointed out by Wedding, this is not a perfect tool and is not a substitute for common sense and process knowledge, because complete reliance on the decision tree may lead to false conclusions .
To determine the kind of variability that might exist at a specific CCP, data of the parameter used to maintain control at this location must be collected and statistically analyzed. For example, the conditions that influence variability at a CCP, like pasteurization temperature for fresh juices, must be
Not CCP CCP
FIGURE 15.2 Critical control point decision tree.
Not CCP CCP
understood and controlled within an acceptable range to ensure that a safe product is manufactured. Processes with a high degree of variability, especially when that variability is not recognized or understood, are more likely to produce unacceptable and possibly hazardous food .
Before data collection begins, the appropriate control chart must be determined for the parameter to be evaluated at the CCP. Two principal categories of control charts are employed in SPC work: variable and attribute. Variable control charts use actual measurements (e.g., temperatures, chlorine concentration, oxidation-reduction potential (ORP), pH values) for charting. Attribute control charts use pass-fail information (metal inclusion, cull fruit presence, foreign objects) for charting. Smith presents a good description of the different types of charts in each category .
To create the chart, individual data, normally arranged into subgroups, are sampled from the process. The average value of the data is then calculated and becomes the centerline of the chart. Using statistical formulae specific for each chart type, upper and lower trial control limits are calculated.
They describe the spread of the process. Finally, the individual (or averaged) measured values are plotted on the control chart. Once the chart is constructed, it presents a picture of the types of variation occurring in the process over the time at which the samples were taken. If one or more plotted points exceed either trial control limit, special cause variation has become a part of the process, forcing it out of statistical control. If a cause can be assigned to each value exceeding the control limit, then it can be discarded from the data and new control limits can be computed from the remaining data. However, if no cause can be found and corrected, then the points cannot be removed from the chart . Once the assignable causes have been eliminated, the revised control charts should be in control. If a process shows only common cause variation present, it is stable, and the process of improvement can begin.
When first implemented, SPC will do a good job of finding areas of high variability (special cause variation) in a process. This results in readily demonstrated points exceeding the control limits. However, as more problems are solved, those remaining will be more subtle in their variation .
When an unusual number of nonrandom points produces a pattern on a control chart, none being beyond the control limits, this signifies that the process is unstable and on the verge of going out of control. While a dozen or more of these patterns may occur in a process, Evans has characterized the five most common ones, as shown in Figure 15.3 : (a) shift — seven or more consecutive points on one side of the center line of a control chart; (b) run — a pattern of seven points consecutively climbing or falling in a control chart; (c) cycling — short repeated patterns of points having alternate high peaks and low valleys on a control chart; (d) instability — unnatural and erratic swings on both sides of the chart over time with points often lying near or on the control limits; and (e) stratification — 14 or more consecutive points hugging the center line on the control chart. When these patterns occur, it is a warning signal that something has gone wrong in the process and immediate action is needed to avoid loss of control.
Was this article helpful?