How do I check and handle data quality in PI System?

Every PI value has an associated quality status that indicates whether the data is good, questionable, or bad. Understanding quality is essential for reliable calculations and reporting.

PI Quality Flags

PI uses OPC-style quality flags:

• Good (192+): data is reliable
• Questionable (64-191): data may be unreliable
• Bad (0-63): data is not usable (substituted, sensor failure, etc.)

Checking Quality in AF Analytics

' Check if a value is good
If BadVal('Temperature') Then
    NoOutput()  ' Don't write bad calculations
Else
    'Temperature' * 1.8 + 32  ' Convert to Fahrenheit
EndIf

' Get percent good over a period
PctGood('Temperature', 't-1h', '*')

Common System Digital States (Bad Quality)

State	Meaning
I/O Timeout	Interface lost connection to data source
Shutdown	PI server was not running
No Data	No value available
Pt Created	Tag was just created
Bad Input	Source reported bad quality
Under Range / Over Range	Value outside sensor range

In Code (AF SDK)

AFValue val = attribute.GetValue();
if (val.IsGood)
{
    double numericValue = val.ValueAsDouble();
    // Process good data
}
else
{
    Console.WriteLine($"Bad quality: {val.Status}");
}

Best Practices

• Always check BadVal() before using values in AF Analytics.
• Use PctGood() in reports to show data reliability.
• Configure stale timeout on interfaces to detect frozen values.
• Don't ignore questionable data — investigate root causes.