Written es before a consonant occurs at the beginning of a very large number of Portuguese words. A quick look into any dictionary will reveal that. The phoneme /s/, being before another consonant, gets the allophonic realisation [S] here, as is normal in the context VsC.
The special thing is, however, that the vowel, which might be a short /i/ or /1/, is in fact completely inaudible, let me venture into saying absent, even in slow speech or song. The vowel drowns in the [S] sound, so these words simply start with [S].
Many years ago, when I thought I already understood all details of the pronunciation of Portuguese, but in hindsight had only discovered the basics, this lead me to believe that a final "s" (pronounced [S]), and an initial es (also pronounced [S]) would merge into a single [S]. So I thought esses esforços © would be pronounced [es1SforsuS].
But this is not so! (but sometimes it is!)
In this situation, the inaudible vowel rears its beautiful head, and turns the final s of "esses" into [z]. This is because the phoneme is between vowels now, and this rule, as you might recall, extends over word boundaries. So the correct pronunciation is: [es1z1SforsuS] ©
In quick speech, the vowels will become very short, and might eventually disappear again, but they leave their traces by influencing the consonants: [eszSforsS] or with a devoiced (but still lax!) [z]: [essSforsS]. Not surprising that this language tends to be difficult to understand for the uninitiated listener, expecting a "normal" European language.
Here is a nice example of a lyrics line,
sung by Camané, in which the phenomenon occurs twice.
This shows that in fact both possibilities (combining final
and initial /S/, or keeping them separate) occur:
Os degraus, quanto mais altos, mais estragadas estão ©, [uZd1grawS kw3~tumajzaltuS majz1Str3gad3St3~w~]. Separate in the first case (mais estragadas), combined in the second (estragadas estão).