Risk assessment methods - James Vesper

there are many different tools and methods that we can use in doing risk assessment in this video we're going to look at three the first is preliminary risk analysis the second failure mode effects analysis and the third fault tree analysis preliminary risk analysis is probably the easiest tool that you can use you're going to use the PRA when you don't have a lot of information it could be at the beginning of a very new project or if you're going to use some other tools later on to dive more deeply into what those risks are there are two different forms of preliminary risk analysis that we'll look at the first is a narrative form and then secondly a spreadsheet or table form the narrative PRA is what you would use if you are doing for example a change control you would look at a particular change that you're going to make and then you'd be taking a look at some of the different areas of interest you'd be asking questions like what could go wrong what things could happen that you would not want to see you'd be asking then about what can you do to prevent those problems from occurring what sort of controls and mitigations would you be putting in place the table based PRA is often frequently used you would use this and then you would add a report at the end of this to summarize what your findings were this is a typical form that will usually see when doing a PRA and with other risk assessment tools one of the first things that you're going to do is ask the risk question your risk question is going to be able to help you focus that risk assessment so in this example what are the risks to transporting vaccines during winter weather when the temperature might be below minus 20 degrees C you're first asking the question what could happen what could go wrong you're putting that information there in that first column the next question that you would ask would be what could be the harm or consequence what would be the impact if something would go wrong here in this case if a vaccine freezes during shipment the vaccine could become inactive the third step is to identify the potential causes of that unwanted event in this case it could be poor controls during shipping or it could be an inadequate shipping container after you identify the potential causes the next step is to identify the likelihood of occurrence how likely is it that this hazard and this harm will occur often you're going to use a rating scale for this some rating scales are low medium or high likelihood other types of rating scales use numbers from 1 to 3 1 to 5 or 1 to 10 in this case we've identified the potential likelihood of occurrence as being 3 for poor controls during shipping and 1 for inadequate shipping containers after the likelihood of occurrence we then identify the severity of the consequence in other words how bad could it get also with this we would use a scale from low medium to high or a number scale 1 2 3 1 through 5 or 1 to 10 now that we have the ratings for likelihood of occurrence and potential impact we can multiply those numbers together and come up with a risk score so in this case we've done that and we have a risk score for that first incident of poor controls during shipping as a 15 the risk score for the second situation inadequate shipping container is 5 what this allows us to do is to say that poor controls during shipping is more significant and needs more attention than the inadequate shipping container another way of showing this information is using a heat map here in this example we have those two incidences that are plotted against a heat map in the heat map you can see different colors the red corresponds to those areas that are extremely high of risk the orange those that are high yellow indicates moderate risk and gray is low risk what you can see is that first incident where we don't have adequate controls that is extremely high we want to focus on that for our treatment process and then the second situation is in the orange which corresponds to a high level of risk now that we've identified those areas where we're going to be putting our resources and putting our attention we can come up with some ways to control that the next column there in step five is where we're putting that information for those controls to treat risk that is to prevent the risk from happening or to protect those things of value so that's the way that we do the preliminary risk analysis we can see that it's a rather simple tool it's very useful and it's very flexible the second method for performing a risk assessment that we'd like to look at is the FMEA our failure mode effects analysis FMEA is probably the most widely used tool that's out there because of its flexibility using FMEA you can take a look at the risks of a design a product a process a system or a particular piece of equipment you can go from the large part of equipment down to the very smallest part a gasket a nut or a bolt so you have a lot of flexibility when you use an FMEA the other aspect of an FMEA is that you're using a spreadsheet you're using a table to collect your information you do need to have quite a bit of information available to you before you perform an FMEA as you can see there are different columns that are used for collecting the information one of the first things that you have to do before performing that FMEA is to designate what you're talking about you have to define your system you have to define your design you're also needing to identify the scope and set borders around which you're focusing on as you're doing the FMEA as you can see in this form that we have your first asking the risk question and in this case what are the risks that the alarms will not work correctly for this particular refrigeration system you can see here in this column that we're looking at the process step or parameter what is supposed to happen here you're identifying success after you've identified success you identify the failure mode in other words what can fail in this example we have the alarm does not send the signal or the alarm sends a no power signal when there actually is power that's the failure mode after that you identify the failure effect what could happen if this particular thing fails in the first situation when the alarm does not send the signal there could be a power loss to the freezer you might not know about that and the contents could be subject to thawing in the other situation when the alarm sends a no power signal when there actually is power you'll get a false alarm and you'll have resources that are expended unnecessarily in responding to that alarm the next column takes a look at the failure mechanism and here you're asking the question what specifically causes this failure mode to occur in the case of the alarm not sending a signal it could be a faulty sensor or it could be that the alarms backup battery fails in the second situation where the alarm sends a no power signal when there actually is power again it could be a faulty sensor as with the other risk assessment tools once you've identified those effects and those mechanisms which are going to do is you're going to rate them with an FMEA we're looking at the likelihood of occurrence again and then the severity of the consequence as with other tools here you will also be using a scale you could use from 1 to 5 or 1 to 10 generally when we're using an FMEA we're using a numerical scale so the results are semi quantitative one thing that's different in the FMEA is you're looking at the detectability of the failure you don't do this all the time with F mes but generally you do and what you can see in this column is how detectable is the failure while something can still be done in other words this is before that unwanted impact actually does occur at this point you now know the likelihood of occurrence the severity of the impact and the detectability of this unwanted event now with these three things you can multiply them together and come up with the risk priority number the risk priority number is very important because you can use this in helping to establish which risks are more significant or more important than the others you're doing this in your risk evaluation step now you can see in this example where we have calculated the risk priority number you see that the most significant risk is rated at 48 that risk corresponds to the category of a major risk the other two are smaller numbers and those correspond to the category of low risk you still should do something with those low risk numbers are those low risk situations because they could occur and you want to be prepared should something unwanted occur so that's the FMEA failure mode effects analysis it's very flexible you need to have considerable information about the system the design the product that you are looking at but you have a lot of flexibility in how you can use it the third method that we're going to look at for performing a risk assessment is the fault tree analysis or FTA fault tree analysis is very different from the others that we've looked at so far in that it's a graphical representation we're not using a table we're not using a type of form what we're doing is we're putting together a causal chain of how this unwanted event could have occurred the other thing that's different with a fault tree analysis is that we're asking a different question for preliminary risk analysis and failure mode effects analysis the question is what if what if this happens what could occur with a fault tree analysis the question is what could cause this unwanted event to happen another thing that's very different with the fault tree analysis is that we're using very different symbols to present different events and as you can see in this diagram these are some of the symbols that we're using what you see here are the different symbols the top one is called the top event and in this case we're wanting to find out what could cause this top event the other symbols that you see are the failure pathways and in this example you can see that there are different shapes that represent these different pathways it's important that you're using these correct symbols when you're performing a fault tree analysis these are highly standardized as you can see in this diagram we have the fault event and then we have another rectangle there that is an intermediate event if you see a diamond-shaped symbol that is an undeveloped event what you're doing here is you're representing it you know that it's something that could happen but for whatever reason you're not interested in taking that farther two other symbols are called gates and gates or logical connectors in one case you could have an or gate that is connecting a intermedia event with other events that follow what or means is that either this or that could cause the event immediately above it with an and gate you need two things at least to occur you need this and that to happen for that fault or that intermediate event at the bottom what you see are circles and those represent basic events you could consider them to be the root cause that you're looking for so by putting this type of a pathway together what you have is this causal tree one pathway that could cause this top event to occur so let's take a look at an example and in this case our top event is what could cause the failure of the refrigerator to maintain the correct temperature and what you see is we have a variety of different situations that could cause that one is that the refrigerator is not Kay people well we're not too interested in that so that's why that is in a diamond shape we have improper storage in the refrigerator we're interested we're going to take that farther we have mechanical failure we have placement in a non controlled environment of that refrigerator again we're not too interested in that or we could have a power failure and as you see we're taking that through a causal chain when you're constructing a fault tree there are a couple of very important rules that you need to follow when you're putting it together the first rule is that there are no gate to gate connections in other words you can't have an or gate directly connecting to a NAND gate that's considered illogical and it cannot occur the other rule as you're doing this you're going to do level by level you're not going through all the way down a causal pathway but you're doing it in sections going across level by level as you can see in this diagram as you are getting ready to perform the fault tree analysis it's important that you understand what should happen what is success and then what you're doing is you're using the success to identify what are the some of the things that could go wrong so let's do a fault tree analysis and see how we can put it together the first thing that you're going to do is to very clearly define what that top event is depending on how you define that top event you might go in two different pathways that could cause that top event so this is what we're concerned about here is that failure of a refrigerator to maintain correct temperatures then what you do is you brainstorm what are some of the things that could cause that this might be based on previous experiences or abbreviated similar equipment so in this case we've identified those things as we said before the refrigerator not capable improper storage mechanical failure placement in a non controlled environment or a power failure then what you're going to do is you're going to focus on what is of most concern to you this is where you're putting those different boxes around those ideas for those that you're not too concerned about but you just want to refer to saying that we considered those you're using the diamonds for those other events you're going to put a rectangle after you've made those two levels now you need to connect them with a logical connector or a gate so in this case what we're saying that each of these events is necessary and sufficient on its own to be able to cause that failure so in this case we're going to use an or gate and connect those to that top event we now continue down to that next level and the level underneath that of identifying more events that it could occur what we're trying to get to are these base events these root causes and those root causes are those things where we've understood that we don't want to go any farther or we cannot go any farther to pull those apart to identify what could cause those to happen so in this model here you can see that we've created three more levels and we've connected them with connectors logical connectors you then can take a look at that fault tree analysis and take a look at how those are situated which are those events that you're most concerned with what are those pathways that you think are most likely to occur if you have data you can use data to support that with the likelihood or probability so that's the fault tree analysis you can use this information on its own it could be standalone or you could take it back to the failure mode effects analysis and that really defines that failure mechanism we've now covered three important tools that we can use in performing a risk assessment we've talked about preliminary risk analysis something that is easy to do very versatile in terms of its formats and how you can present the information we've talked about the FMEA or failure mode effects analysis probably one of the most widely used tools in the pharmaceutical industry it's very popular because of its flexibility and how you can go from very large to very small in doing your analysis third tool that we took a look at was fault tree analysis a graphical tool that takes a look at what could cause a potential unwanted situation with those three tools you can use them to understand risk and to identify better ways to treat those risks to prevent them from happening and should they occur protect those things of value

Channel: umit kartoglu

Views: 66,502

Rating: 4.9385562 out of 5

Keywords: Risk assessment methods, PRA, EPELA, James Vesper, authentic e-learning, e-pharmaceutical cold chain management

Id: olMKwMzEcyU

Length: 18min 23sec (1103 seconds)

Published: Fri Jan 10 2014