White Paper - Leveraging Sensory Testing to Optimize Processing ...

FOOD FOR THOUGHT Topical Insights from our Subject Matter Experts

LEVERAGING SENSORY TESTING TO OPTIMIZE PROCESSING PARAMETERS

The NFL White Paper Series Volume 4, July 2012

Overview When new food products are developed, many decisions need to be made in addition to the final recipe, including choosing processing parameters. While process methods are sometimes dictated by what equipment is available at a given plant, there are often many options available and designed trials can be set up to test the potential options. The obvious primary goal for these tests is to assure safety; all product needs to be heated sufficiently to achieve commercial sterility. A secondary goal of these trials can be to choose a processing method that optimizes the flavor and texture profile of the finished product; we want to create a product that is liked by consumers. Processing time and associated costs also weigh in on the final processing method decision. The NFL designed a set of retort processing trials for a canned Queso Pasta product (Figure 1) in order to compare processing times and examine the sensory differences across the products. Samples were processed using seven different retorting modes in an Allpax R&D Retort. Trained sensory panelists then evaluated the products using a Narrative Descriptive method to help choose among the available options.

Figure 1: Queso Pasta

Allpax Retort Test Design The NFL’s Product Innovation team developed a recipe for Queso Pasta (penne pasta in a spicy cheese sauce with vegetables) to be used in this experiment as an example of a low acid canned food (Figure 2). Ingredients were packed into 15 ounce (300 x 407) cans for processing. Seven different water spray retort agitation modes were chosen for this study: still, three speeds of end over end rotation, and three speeds of gentle motion (back and forth movement) (Table 1). These methods were chosen because they are readily available in the food processing industry. The NFL’s Allpax R&D Retort was used for all of the runs. Eight cans in each run were equipped with thermocouples located in their geometric centers in order to monitor product temperatures. Samples were retorted until the slowest heating can achieved commercial sterility (F0 = 6.0 minutes) and then cooled below 130°F. Figure 2: Queso Pasta Recipe

Table 1: Allpax Retort Processing No Motion: Still (Reference) End Over End: 5 rotations/minute 10 rotations/minute 15 rotations/minute Gentle Motion: 30 strokes/minute 60 strokes/minute Shaka: 90 strokes/minute

Food for Thought/White Paper Series – Volume 4 (Leveraging Sensory to Optimize Processing Parameters)

Page 2 of 5

Narrative Sensory Testing At The National Food Lab, we typically use Narrative Descriptive Analysis when our clients need indepth written descriptions of product sensory characteristics but do not need quantitative data or statistical analysis. This method provides rapid feedback comparing the sensory properties of multiple samples. An NFL panel leader and three members of The NFL’s trained descriptive panel participated in these evaluations. The panelists independently evaluated each sample and described the major appearance, aroma, flavor, and texture characteristics of the Still Cook (Reference) sample. They then indicated the specific sensory attributes that differed in the Variant samples and to what degree the Variant samples differed from the Reference sample in the specific attributes noted. In addition, they made note if a Variant sample showed consumer relevant downsides compared to the Reference sample. All of the information collected was based upon a consensus of opinion reached though group discussions under the guidance of the panel leader. We have found the Narrative Descriptive output to be easily understood by all audiences.

Results Meaningful differences in processing time, appearance, aroma, flavor, and texture were found among the experimental samples. Processing times are shown in Figure 3 and Sensory results for select products are shown in Table 2.

Processing Time

Figure 3: Total Processing Time 140 Cooling 120

Cooking

100 Time [min]

The various agitation methods yielded large differences in processing time. For the treatment levels tested here, processing time decreased as rotation speed or back-and-forth motion speed increased. The total processing time (excluding come up time) ranged from 129 minutes for the Still Cook down to 48 minutes for Shaka at 90 strokes/minute (Figure 3).

80 60 40 20 0 Still 

EoE 5  EoE 10  EoE 15  GM 30  GM 60  Shaka  rpm rpm rpm spm spm 90 spm

Visual Characteristics

When cans of the Still Cook (Reference) sample were opened, the sample was not evenly distributed; the vegetables were found predominantly in the top half of the can while the cheese and pasta were compressed and stuck to the can walls on the bottom half of the can. In addition, the cheese sauce was slightly darker near the top of the can and there was some visible oil separation. In contrast, the End Over End rotation treatments and the Shaka treatment all showed even distribution of the pasta and vegetables. The product was not stuck to the sides of the cans and

Food for Thought/White Paper Series – Volume 4 (Leveraging Sensory to Optimize Processing Parameters)

Page 3 of 5

there was no visible oil separation. This makes sense; the Still Cook (Reference) sample had the longest heating time and was not moved during processing, so the product settled unevenly and cooked onto the sides of the can, while the End Over End rotation treatments and the Shaka treatment were mixed by the retort’s motion during cooking. The Gentle Motion treatments were similar to the Still Cook product in sample distribution; the product was not moved enough to mix the sample thoroughly during these treatments. The Shaka sample’s sauce had a moderately more grainy appearance than the Still Cook sample’s sauce. This is likely due to a slight breakdown of the pasta during the 90 strokes/minute back and forth motion retort process. The other Variant samples showed much less graininess than the Shaka sample.

Flavor and Texture Characteristics

The flavors of the Still Cook (Reference) sample were characterized as moderate nacho cheese, moderate chemical heat, moderate savory character, moderate salt, slight to moderate bell pepper and chili pepper, slight tomato, slight onion, and slight pasta/starch. The pasta and vegetables were soft and the cheese sauce was moderately thick. The two Gentle Motion treatments (30 strokes/minute and 60 strokes/minute back and forth) were the most similar to the Still Cook Reference sample in aroma, flavor, and texture. The rest of the Variants showed less intense nacho cheese and bell pepper notes, more prominent savory notes, and thinner sauce texture to varying degrees. No off-notes were found with the exception of the Shaka treatment sample; its savory character was less fresh and more chicken-broth like than the other samples. Table 2: Sensory Narrative Descriptive Results for Select Products Reference Retort Process Appearance in Can

Appearance in Bowl

Aroma

Flavor

Texture Diff. from Reference Consumer Relevance

Test Products - Deviations from Reference

Water spray, No motion Some slightly oily liquid on top, golden tan cheese sauce color on top of can, vegetables mostly on top, cheese and pasta compressed and stuck to sides of can on bottom half of can Moderately thick creamy/smooth dark golden yellow cheese sauce, 1inch tubes of diagonally cut ribbed pasta Moderate nacho cheese, moderate bell pepper, moderate savory aromatic, slight tomato, very slight onion, very slight starch Moderate nacho cheese, moderate chemical heat, moderate savory, moderate salt, slight to moderate bell pepper and chili pepper, slight tomato, slight onion, slight starch Soft pasta, moderately thick and very slightly grainy cheese sauce, soft vegetables

Water spray, 15 rpm end over end No liquid on surface, slightly lighter cheese sauce color on top of can, cheese and vegetables evenly distributed, product not stuck to sides of can, slightly lighter sauce on top of can

Water spray, Shaka, 90 strokes/min No liquid on surface, cheese and vegetables evenly distributed, product not stuck to sides of can, moderately grainy appearance

Slightly lighter sauce color, slightly more grainy sauce

Moderately more grainy sauce, slightly lighter sauce

Slightly less nacho cheese, slightly less bell pepper

Slightly to moderately less nacho cheese, slight to moderate chicken broth character (less fresh savory character)

Slightly less nacho cheese, moderate savory is slightly more prominent, slightly more pasta/starch

Moderately less nacho cheese, moderate chicken broth savory character, moderately less bell pepper

Slightly thinner sauce, very slightly firmer pasta

Slightly to moderately thinner sauce, slightly firmer pasta

REFERENCE

Slight Difference

Moderate Difference

REFERENCE

Acceptable Alternative

Moderate Downside

Food for Thought/White Paper Series – Volume 4 (Leveraging Sensory to Optimize Processing Parameters)

Page 4 of 5

Product Decision In this trial, the ideal processing method would minimize processing time while providing a high quality sensory profile. The two fastest methods, End Over End 15 rotations/minute and Shaka 90 strokes/minute, resulted in roughly a 60% reduction in total processing time, boosting potential production capacity significantly and lowering utility costs considerably on a per can basis. If the final processing decision were made on speed alone, either of these options would be a reasonable choice. However, these two rapid methods have very different sensory profiles (Table 2). The Shaka sample showed noticeable consumer relevant downsides; its appearance was noticeably more grainy due to pasta breakdown, its nacho cheese flavor was moderately less intense, and its savory notes were less fresh tasting than the Reference sample. In contrast, the End Over End 15 rotations/minute sample showed slight differences from the Reference sample, but none of these differences were considered downsides. In this experiment, descriptive sensory testing makes the End Over End 15 rotations/minute treatment the clear choice. The NFL’s Consumer Research team could then test the chosen product with its target consumers to make sure it is well liked and meets the client’s standards for product launch.

Flexible Test Design The National Food Lab’s integrated services make it easy to custom design processing trials to meet the needs of any project. Our Product Innovation, Commercialization, Sensory Research, Consumer Testing, and Shelf Life groups work seamlessly together to design and execute processing experiments, making us a one-stop shop for complex study execution. These types of trials can be conducted for any combination of processing methods or package types. The NFL’s Commercialization group has equipment to support retort, aseptic, and hot fill processing methods in our state of the art pilot plant. Our Allpax Multimode R&D Retort can process in still steam, continuous rotary steam, water cascade, full or partial water immersion, endover-end rotation, gentle motion shaking, and Shaka modes; this flexibility allows us to quickly design and test processing options without the cost of shutting down production lines. As new food products are developed, attention should be given to choosing commercially feasible processing methods that will result in high quality product with low processing costs. Conducting processing and sensory trials such as those summarized above can help bring higher quality products to market. About The NFL The National Food Laboratory is a food and beverage consulting and testing firm providing creative, practical and science-based solutions for the following areas: Food Safety and Quality; Product and Process Development; and Sensory and Consumer Research. We create value for our clients by enabling them to develop commercially safe, high quality and great tasting foods and beverages. For more information about The National Food Laboratory, please visit us at www.TheNFL.com. For more information please contact: Terry Berman at [email protected], 925.556.4813 or Dawn Chapman at [email protected], 925.551.4243 Food for Thought/White Paper Series – Volume 4 (Leveraging Sensory to Optimize Processing Parameters)

Page 5 of 5