One way you can reduce your query rows is to apply the filters you are testing for in the first for loop.

However, you want to avoid field != null queries as they are very expensive and result in a full table scan.

One thing you can do is make three calculation fields on your contact object

eg Work_Email_Not_Blank__c and define it something like:

if (isBlank(con.npe01__WorkEmail__c),FALSE,TRUE)

Then you can change your top query to something like:

SELECT Id,Email,npe01__WorkEmail__c,npe01__HomeEmail__c 
FROM Contact
WHERE Work_Email_Not_Blank__c = true 
  AND Email_Not_Blank__c
  AND Home_Email_Not_Blank__c

One way you can reduce your query rows is to apply the filters you are testing for in the first for loop.

However, you want to avoid field != null queries as they are very expensive and result in a full table scan.

One thing you can do is make three calculation fields on your contact object

eg Work_Email_Not_Blank__c and define it something like:

if (isBlank(con.npe01__WorkEmail__c),FALSE,TRUE)

THEN you need to use a workflow or process to copy these fields into static versions eg Work_Email_Not_Blank__c -> Work_Email_Not_Blank_Static__c

Then you can change your top query to something like:

SELECT Id,Email,npe01__WorkEmail__c,npe01__HomeEmail__c 
FROM Contact
WHERE Work_Email_Not_Blank_Static__c = true 
  AND Work_Email_Not_Blank_Static__c
  AND Work_Email_Not_Blank_Static__c